Extreme Cloud/Cluster Administration Toolkit¶
xCAT stands for Extreme Cloud/Cluster Administration Toolkit.
xCAT offers complete management of clouds, clusters, HPC, grids, datacenters, renderfarms, online gaming infrastructure, and whatever tomorrows next buzzword may be.
- xCAT enables the administrator to:
- Discover the hardware servers
- Execute remote system management
- Provision operating systems on physical or virtual machines
- Provision machines in Diskful (stateful) and Diskless (stateless)
- Install and configure user applications
- Parallel system management
- Integrate xCAT in Cloud
You’ve reached xCAT documentation site, The main page product page is http://xcat.org
xCAT is an open source project hosted on GitHub. Go to GitHub to view the source, open issues, ask questions, and particpate in the project.
Enjoy!
Table of Contents¶
Admin Guide¶
The admin guide is intended to help with learning how to manage a cluster using xCAT with the following major sections:
- Basic Concepts Introduces some of the basic concepts in xCAT.
- Manage Cluster Describes managing clusters under xCAT. The management procedures are organized based on the hardware type since management may vary depending on the hardware architecture.
- Reference xCAT reference sections.
References¶
xCAT Man Pages¶
These man pages are auto generated from .pod files to .rst files using the create_man_pages.py script under xcat-core
man1¶
addkit.1¶
addkit - Install a kit on the xCAT management node
addkit [-? | -h | --help] [-v | --version]
addkit [-i | --inspection] kitlist
addkit [-V | --verbose] [-p | --path path] kitlist
The addkit command install a kit on the xCAT management node from a kit tarfile or directory, creating xCAT database definitions for kit, kitrepo, kitcomponent.
Note: The xCAT support for Kits is only available for Linux operating systems.
-h|--help
Display usage message.
-V|--verbose
Verbose mode.
-v|--version
Command version.
-i|--inspection
Show the summary of the given kits
-p|--path <path>
The destination directory to which the contents of the kit tarfiles and/or kit deploy dirs will be copied. When this option is not specified, the default destination directory will be formed from the installdir site attribute with ./kits subdirectory.
kitlist
a comma delimited list of kit_tarball_files and kit_deploy_dirs that are to be added to the xCAT cluster. Each entry can be an absolute or relative path. For kit_tarball_files, these must be valid kits tarfiles added. For kit_deploy_dirs, these must be fully populated directory structures that are identical to the contents of an expanded kit_tarball_file.
- To add two kits from tarball files.
addkit kit-test1.tar.bz2,kit-test2.tar.bz2
Output is similar to:
Kit /kit/kit-test1.tar.bz2,/kit/kit-test2.tar.bz2 was successfully added.
- To add two kits from directories.
addkit kit-test1,kit-test2
Output is similar to:
Kit /kit/kit-test1,/kit/kit-test2 was successfully added.
- To add a kit from tarball file to /install/test directory.
addkit -p /install/test kit-test1.tar.bz2
Output is similar to:
Kit /kit/kit-test1.tar.bz2 was successfully added.
- To read the general infomration of the kit, without adding the kits to xCAT DB
addkit -i kit-test1.tar.bz2
Output is similar to:
kitname=xlc-12.1.0.0-Linux description=XLC12 for Linux version=12.1.0.0 ostype=Linux
lskit(1)|lskit.1, rmkit(1)|rmkit.1, addkitcomp(1)|addkitcomp.1, rmkitcomp(1)|rmkitcomp.1, chkkitcomp(1)|chkkitcomp.1
addkitcomp.1¶
addkitcomp - Assign Kit components to an xCAT osimage.
addkitcomp [-? | -h | --help] [-v | --version]
addkitcomp [-V | --verbose] [-a | --adddeps] [-f | --force] [-n | --noupgrade] [--noscripts] -i osimage kitcompname_list
The addkitcomp command will assign kit components to an xCAT osimage. The kit component meta rpm, package rpm and deploy parameters will be added to osimage’s otherpkg.pkglist and postbootscripts will be added to osimages’s postbootscripts attribute.
Note: The xCAT support for Kits is only available for Linux operating systems.
-a|--adddeps
Assign kitcomponent dependencies to the osimage.
-h|--help
Display usage message.
-V|--verbose
Verbose mode.
-v|--version
Command version.
-f|--force
Add kit component to osimage even if there is a mismatch in OS, version, arch, serverrole, or kitcompdeps
-i osimage
The osimage name that the kit component is assigning to.
-n|--noupgrade
Allow multiple versions of kitcomponent to be installed into the osimage, instead of kitcomponent upgrade
--noscripts
Do not add kitcomponent’s postbootscripts to osimage
kitcompname_list
A comma-delimited list of valid full kit component names or kit component basenames that are to be added to the osimage.
- To assign a kit component to osimage
addkitcomp -i rhels6.2-ppc64-netboot-compute comp-test1-1.0-1-rhels-6.2-ppc64
Output is similar to:
Assigning kit component comp-test1-1.0-1-rhels-6.2-ppc64 to osimage rhels6.2-ppc64-netboot-compute Kit components comp-test1-1.0-1-rhels-6.2-ppc64 were added to osimage rhels6.2-ppc64-netboot-compute successfully
- To assign a kit component to osimage with its dependency.
addkitcomp -a -i rhels6.2-ppc64-netboot-compute comp-test2-1.0-1-rhels-6.2-ppc64
Output is similar to:
Assigning kit component comp-test1-1.0-1-rhels-6.0-ppc64 to osimage rhels6.2-ppc64-netboot-compute Assigning kit component comp-test2-1.0-1-rhels-6.2-ppc64 to osimage rhels6.2-ppc64-netboot-compute Kit components comp-test1-1.0-1-rhels-6.0-ppc64,comp-test2-1.0-1-rhels-6.2-ppc64 were added to osimage rhels6.2-ppc64-netboot-compute successfully
- To assign a kit component to osimage with incompatable osarch, osversion or ostype.
addkitcomp -f -i rhels6.2-ppc64-netboot-compute comp-test1-1.0-1-rhels-6.2-ppc64
Output is similar to:
Assigning kit component comp-test1-1.0-1-rhels-6.2-ppc64 to osimage rhels6.2-ppc64-netboot-compute Kit components comp-test1-1.0-1-rhels-6.2-ppc64 were added to osimage rhels6.2-ppc64-netboot-compute successfully
- To assign a new version of kit component to osimage without upgrade.
addkitcomp -n -i rhels6.2-ppc64-netboot-compute comp-test2-1.0-1-rhels-6.2-ppc64
Output is similar to:
Assigning kit component comp-test1-1.0-1-rhels-6.0-ppc64 to osimage rhels6.2-ppc64-netboot-compute Assigning kit component comp-test2-1.0-1-rhels-6.2-ppc64 to osimage rhels6.2-ppc64-netboot-compute Kit components comp-test2-1.0-1-rhels-6.2-ppc64 were added to osimage rhels6.2-ppc64-netboot-compute successfully
The result will be: lsdef -t osimage rhels6.2-ppc64-netboot-compute -i kitcomponents Object name: rhels6.2-ppc64-netboot-compute kitcomponents=comp-test2-1.0-0-rhels-6.2-ppc64,comp-test2-1.0-1-rhels-6.2-ppc64
lskit(1)|lskit.1, addkit(1)|addkit.1, rmkit(1)|rmkit.1, rmkitcomp(1)|rmkitcomp.1, chkkitcomp(1)|chkkitcomp.1
buildkit.1¶
buildkit - Used to build a software product Kit which may be used to install software in an xCAT cluster.
buildkit [-? | -h | --help] [-v | --version]
To build a new Kit
buildkit [-V | --verbose] subcommand [kit_name] [repo_name|all] [-l | --kitloc kit_location]
To add packages to an existing Kit.
buildkit [-V | --verbose] addpkgs kit_tarfile [-p | --pkgdir package_directory_list] [-k | --kitversion version] [-r | --kitrelease release] [-l | --kitloc kit_location]
The buildkit command provides a collection of utilities that may be used to package a software product as a Kit tarfile that can be used to install software on the nodes of an xCAT cluster. A Kit contains the product software packages, configuration and control information, and install and customization scripts.
Note: The xCAT support for Kits is only available for Linux operating systems.
You will need to run the buildkit command several times with different subcommands to step through the process of building a kit:
By default the buildkit subcommands will operate in the current working directory, (ie. look for files, create directories etc.). You could specify a different location by using the “-l | --kitloc kit_location” option.
The kit_location is the full path name of the directory that contains the kit files. You would use the same location value for all the buildkit subcommands.
For example, to create a new kit named “prodkit” in the directory /home/mykits/ either run:
If no location is provided then the command will create a subdirectory called “prodkit” in the current directory “/home/mykits” and the new kit files will be created there.
cd /home/mykits
buildkit create prodkit
or
If a location is provided then the Kit files will be created there. Note that the Kit name does not necessarily have to be the directory name where the kit files are located.
buidkit create prodkit -l /home/mykits/prodkit
In both cases the /home/mykits/prodkit directory is created and the inital files for the kit are created in that directory.
The following example illustrates the basic process for building a new Kit. In this example we are building a Kit named “mytstkit”.
Change to the directory where you wish to create the Kit.
Create a template directory for your kit:
buildkit create mytstkit
Change directory to the new “mytstkit” subdirectory that was just created.
cd mytstkit
Edit the buildkit configuration file for your kit:
vi buildkit.conf
(See xCAT Kit documentation for details.)
Create all required files, scripts, plugins, and packages for your kit.
Validate your kit build configuration and fix any errors that are reported:
buildkit chkconfig
List the repos defined in your buildkit configuration file:
buildkit listrepo
For each repo name listed, build the repository. Note that if you need to build repositories for OS distributions, versions, or architectures that do not match the current system, you may need to copy your kit template directory to an appropriate server to build that repository, and then copy the results back to your main build server. For example, to build a repo named “rhels6.3” you would run the following command.
buildkit buildrepo rhels6.3
or, you can build all of the repos at one time if there are no OS or architecture dependencies for kitcomponent package builds or kitpackage builds:
buildkit buildrepo all
Build the kit tar file:
buildkit buildtar
-h |--help
Display usage message.
-k|--kitversion version
Product version.
-l|--kitloc kit_location
The directory location of the Kit files.
-p|--pkgdir package_directory_list
A comma-separated list of directory locations for product RPMs.
-r|--kitrelease release
Product release.
-V |--verbose
Verbose mode.
-v|--version
Command version.
create kit_basename
Creates a new kit build directory structure for kit kit_basename using the location specified on the command line or the current directory. The sample kit files from /opt/xcat/share/xcat/kits/kit_template are copied over, and the buildkit.conf file is modified for the specified kit_basename.
chkconfig
Reads the buildkit.conf file, verifies that the file syntax is correct and that all specified files exist.
listrepo
Reads the buildkit.conf file, lists all Kit package repositories listed in the file, and reports the build status for each repository.
buildrepo {repo_name | all}
Reads the buildkit.conf file, and builds the specified Kit package repository. The built packages are placed in the directory <kit_location>/build/kit_repodir/repo_name. If all is specified, all kit repositories are built.
cleanrepo {repo_name | all}
Reads the buildkit.conf file, and deletes all the package files and package meta data files from the <kit_location>/build/kit_repodir/repo_name directory. If all is specified, all kit repository files are deleted.
buildtar
Reads the buildkit.conf file, validates that all kit repositories have been built, and builds the Kit tar file <kit_location>/kitname.tar.bz2.
cleantar
Reads the <kit_location>/buildkit.conf file and deletes the following:
- Kit tar files matching <kit_location>/kit_name\.tar.bz2*.
- <kit_location>/build/kit_name
- <kit_location>/rpmbuild
- <kit_location>/tmp
- <kit_location>/debbuild
Caution: Make sure you back up any tar files you would like to keep before running this subcommand.
cleanall
Equivalent to running buildkit cleanrepo all and buildkit cleantar.
addpkgs
kit_tarfile {-p | --pkgdir package_directory_list} [-k | --kitversion version] [-r | --kitrelease release]
Add product package rpms to a previously built kit tar file. This is used for partial product kits that are built and shipped separately from the product packages, and are identified with a kit_tarfile name of kitname.NEED_PRODUCT_PKGS.tar.bz2. Optionally, change the kit release and version values when building the new kit tarfile. If kitcomponent version and/or release values are defaulted to the kit values, those will also be changed and new kitcomponent rpms will be built. If kit or kitcomponent scripts, plugins, or other files specify name, release, or version substitution strings, these will all be replaced with the new values when built into the new complete kit tarfile kit_location/new_kitname.tar.bz2.
To create the sample kit shipped with the xCAT-buildkit rpm on a RHELS 6.3 server and naming it mykit, run the following commands:
cd /home/myuserid/kits
buildkit create mykit
cd mykit
vi buildkit.conf
buildkit chkconfig
buildkit listrepo
buildkit buildtar
To clean up a kit repository directory after build failures on a RHELS 6.3 server to prepare for a new kit repository build, run:
buildkit cleanrepo rhels6.3
To clean up all kit build files, including a previously built kit tar file, run
buildkit cleanall
To create a kit named “tstkit” located in /home/foobar/tstkit instead of the current working directory.
buildkit create tstkit -l /home/foobar/tstkit
/opt/xcat/bin/buildkit
/opt/xcat/share/xcat/kits/kit_template
/opt/xcat/share/xcat/kits/kitcomponent.spec.template
<kit location>/buildkit.conf
<kit location>/build/kitname/kit.conf
<kit location>/kitname.tar.bz2
addkit(1), lskit(1), rmkit(1), addkitcomp(1), rmkitcomp(1), chkkitcomp(1)
cfgve.1¶
cfgve - Configure the elements for a virtual environment.
cfgve -t dc -m manager -o object [-c -k nfs|localfs | -r]
cfgve -t cl -m manager -o object [-c -p cpu type| -r -f]
cfgve -t sd -m manager -o object [-c | -g | -s| -a | -b | -r -f]
cfgve -t nw -m manager -o object [-c -d data center -nvlan ID | -a -l cluster | -b | -r]
cfgve -t tpl -m manager -o object [-r]
The cfgve command can be used to configure a virtual environment for ‘Storage Domain’, ‘Network’ and ‘Template’ objects.
The mandatory parameter -m manager is used to specify the address of the manager of virtual environment. xCAT needs it to access the RHEV manager.
The mandatory parameter -t type is used to specify the type of the target object.
Basically, cfgve command supports five types of object: dc, cl, sd, nw and tpl.
dc - The create and remove operations are supported.
cl - The create and remove operations are supported.
sd - The create, attach, detach, activate, deactivate and remove operations are supported.
nw - The create, attach, detach and remove operations are supported.
tpl - The remove operation is supported.
The mandatory parameter -o object is used to specify which object to configure.
-a To attach the target object.
-b To detach the target object.
-c To create the target object.
For creating of Storage Domain, the target storage domain will be created first, then attached to data center and activated.
The parameters that used to create the storage domain are gotten from ‘virtsd’ table. The detail parameters in the virtsd table:
virtsd.node - The name of the storage domain.
virtsd.sdtype - The type of storage domain. Valid value: data, iso, export. Default value is ‘data’.
virtsd.stype - The storage type. “nfs” or “localfs”.
virtsd.location - The location of the storage. nfs: Format: [nfsserver:nfspath]. The NFS export directory must be configured for read write access and must be owned by vdsm:kvm. localfs: “/data/images/rhev” is set by default.
virtsd.host - A host must be specified for a storage doamin as SPM (Storage Pool Manager) when initialize the storage domain. The role of SPM may be migrated to other host by rhev-m during the running of the datacenter (For example, when the current SPM encountered issue or going to maintenance status.
virtsd.datacenter - The storage will be attached to. ‘Default’ data center is the default value.
-d data center
The name of data center.
Specify the ‘Data Center’ that will be used for the object to be attached to. It is used by <nw> type.
-f It can be used with -r to remove the target object by force.
For removing of Storage Domain, if -f is specified, the storage domain will be deactivated and detached from data center before the removing.
-g To activate the target object.
-h Display usage message.
-k storage type
To specify the type of the storage type when creating the data center.
Supported type: nfs; localfs.
-l cluster
Specify the cluster for the network to attach to.
-m manager
Specify the manager of the virtual environment.
For RHEV, the FQDN (Fully Qualified Domain Name) of the rhev manager have to be specified.
-n vlan ID
To specify the vlan number when creating a network.
-o object
The name of the target object.
-p cpu type
To specify the cpu type when creating the cluster. Intel Penryn Family is default type.
Supported type: Intel Conroe Family, Intel Penryn Family, Intel Nehalem Family, Intel Westmere Family, AMD Opteron G1, AMD Opteron G2, AMD Opteron G3
-r To remove the target object.
For removing of Storage Domain, the storage space will be formatted after removing.
-s To deactivate the target object.
-t type
Specify the type of the target object.
- Supported types:
- dc - Data Center cl - Cluster sd - Storage Domain nw - Network tpl - Template
- To create the Storage Domain ‘sd1’, enter:
cfgve -t sd -m <FQDN of rhev manager> -o sd1 -c
- To deactivate the Storage Domain ‘sd1’ from data center, enter:
cfgve -t sd -m <FQDN of rhev manager> -o sd1 -s
- To remove the Storage Domain ‘sd1’, enter:
cfgve -t sd -m <FQDN of rhev manager> -o sd1 -r
- To create the network ‘nw1’, enter:
cfgve -t nw -m <FQDN of rhev manager> -o nw1 -c
- To remove the template ‘tpl01’, enter:
cfgve -t tpl -m <FQDN of rhev manager> -o tpl01 -r
/opt/xcat/bin/cfgve
lsve(1)|lsve.1
cfm2xcat.1¶
cfm2xcat - Migrates the CFM setup in CSM to the xdcp rsync setup in xCAT.
cfm2xcat [-i path of the CFM distribution files generated *] [*-o** *path of the xdcp rsync files generated from the CFM distribution files *]
cfm2xcat [-h]
Copy the cfm2xcat command to the CSM Management Server. Run the command, indicating where you want your files saved with the -i and -o flags. They can be in the same directory. The cfm2xcat command will run cfmupdatenode -a, saving the generated CFM distribution files in the directory indicates with (-i). From those distribution files, it will generate xdcp rsync input files (-F option on xdcp) in the directory indicated by ( -o). Check the rsync files generated. There will be a file generated (rsyncfiles) from the input -o option on the command, and the same file with a (.nr) extension generated for each different noderange that will used to sync files based on your CFM setup in CSM. The rsyncfiles will contain the rsync file list. The rsyncfiles.nr will contain the noderange. If multiple noderanges then the file name (rsyncfiles) will be appended with a number.
-h Display usage message.
-i Path of the CFM distribution files generated from the cfmupdatenode -a command.
-o Path of the xdcp rsync input file generated from the CFM distribution files.
- To build xCAT rsync files to use with xdcp -F , enter on the CSM Management Server, make sure the path exists:
cfm2xcat -i /tmp/cfm/cfmdistfiles -o /tmp/cfm/rsyncfiles
- To use the file on the xCAT Management Node copy to /tmp/cfm on the xCAT MN:
xdcp ^/tmp/cfm/rsyncfiles.nr -F /tmp/cfm/rsyncfilesxdcp ^/tmp/cfm/rsyncfiles.nr1 -F /tmp/cfm/rsyncfiles1xdcp ^/tmp/cfm/rsyncfiles.nr2 -F /tmp/cfm/rsyncfiles2
/opt/xcat/share/xcat/tools/cfm2xcat
chdef.1¶
chdef - Change xCAT data object definitions.
chdef [-h | --help] [-t object-types]
chdef [-t object-types] [-o object-names] [-n new-name] [node]
chdef [-V | --verbose] [-t object-types] [-o object-names] [-d | --dynamic] [-p | --plus] [-m | --minus] [-z | --stanza] [[-w attr==val] [-w attr=~val] ...] [noderange] [attr=val [attr=val...]]
This command is used to change xCAT object definitions which are stored in the xCAT database. The default is to replace any existing attribute value with the one specified on the command line. The command will also create a new definition if one doesn’t exist.
This command also can be used to change the xCAT object name to a new name. Note: the site,monitoring types can NOT be supported.
attr=val [attr=val ...]
Specifies one or more “attribute equals value” pairs, separated by spaces. Attr=val pairs must be specified last on the command line. Use the help option to get a list of valid attributes for each object type.
-d|--dynamic
Use the dynamic option to change dynamic node groups definition. This option must be used with -w option.
-h|--help
Display usage message.
-m|--minus
If the value of the attribute is a list then this option may be used to remove one or more items from the list.
-n new-name
Change the current object name to the new-name which is specified by the -n option. Objects of type site and monitoring cannot be renamed with the -n option. Note: For the -n option, only one node can be specified. For some special nodes such as fsp, bpa, frame, cec etc., their name is referenced in their own hcp attribute, or the hcp attribute of other nodes. If you use -n option, you must manually change all hcp attributes that refer to this name.
noderange
A set of comma delimited node names and/or group names. (must be the first parameter) See the “noderange” man page for details on supported formats.
-o object-names
A set of comma delimited object names.
-p|--plus
This option will add the specified values to the existing value of the attribute. It will create a comma-separated list of values.
-t object-types
A set of comma delimited object types. Use the help option to get a list of valid object types.
-V|--verbose
Verbose mode.
-w attr==val -w attr=~val ...
Use one or multiple -w flags to specify the selection string that can be used to select objects. The operators ==, !=, =~ and !~ are available. Use the help option to get a list of valid attributes for each object type.
- Operator descriptions:
- == Select nodes where the attribute value is exactly this value. != Select nodes where the attribute value is not this specific value. =~ Select nodes where the attribute value matches this regular expression. !~ Select nodes where the attribute value does not match this regular expression.
Note: the operator !~ will be parsed by shell, if you want to use !~ in the selection string, use single quote instead. For example:-w ‘mgt!~ipmi’.
-z|--stanza
Indicates that the file being piped to the command is in stanza format. See the xcatstanzafile man page for details on using xCAT stanza files.
-u
Fill in the attributes such as template file, pkglist file and otherpkglist file of osimage object based on the specified parameters. It will search “/install/custom/” directory first, and then “/opt/xcat/share/”.
Note: this option only works for objtype osimage.
To change a site definition.
chdef -t site -o clustersite installdir=/xcatinstall
To change a basic node definition.
chdef -t node -o node01 groups="all,aix"
(The group definitions are also created if they don’t already exist.)
To add another group to the “groups” attribute in the previous example.
chdef -p -t node -o node01 groups="compute"
To remove the “all” group from the “groups” attribute in the previous example.
chdef -m -t node -o node01 groups="all"
To replace the current “groups” attribute value of “node01”.
chdef -t node -o node01 groups="linux"
To add “node01” to the “members” attribute of a group definition called “LinuxNodes”.
chdef -p -t group -o LinuxNodes members="node01"
To update a set of definitions based on information contained in the stanza file mystanzafile.
cat mystanzafile | chdef -z
To update a dynamic node group definition to add the cons=hmc wherevals pair.
chdef -t group -o dyngrp -d -p -w cons==hmc
To change the node object name from node1 to node2.
chdef -t node -o node1 -n node2
To change the node hwtype, this command will change the value of ppc.nodetype.
chdef -t node -o node1 hwtype=lpar
To change the policy table for policy number 7.0 for admin1
chdef -t policy -o 7.0 name=admin1 rule=allow
To change the node nic attributes
chdef -t node -o cn1 nicips.eth0="1.1.1.1|1.2.1.1" nicnetworks.eth0="net1|net2" nictypes.eth0="Ethernet"
To update an osimage definition.
chdef redhat6img -u provmethod=install
$XCATROOT/bin/chdef
(The XCATROOT environment variable is set when xCAT is installed. The default value is “/opt/xcat”.)
This command is part of the xCAT software product.
mkdef(1)|mkdef.1, lsdef(1)|lsdef.1, rmdef(1)|rmdef.1, xcatstanzafile(5)|xcatstanzafile.5
chhypervisor.1¶
chhypervisor - Configure the virtualization hosts.
RHEV specific :
chhypervisor noderange [-a]
chhypervisor noderange [-n]
chhypervisor noderange [-p]
chhypervisor noderange [-e]
chhypervisor noderange [-d]
zVM specific :
chhypervisor noderange [--adddisk2pool function region volume group]
chhypervisor noderange [--addscsi device_number device_path option persist]
chhypervisor noderange [--addvlan name owner type transport]
chhypervisor noderange [--addvswitch name osa_dev_addr osa_exp_adapter controller connect (0, 1, or 2) memory_queue router transport vlan_id port_type update gvrp native_vlan]
chhypervisor noderange [--addzfcp2pool pool status wwpn lun size owner]
chhypervisor noderange [--removediskfrompool function region group]
chhypervisor noderange [--removescsi device_number persist (YES or NO)]
chhypervisor noderange [--removevlan name owner]
chhypervisor noderange [--removevswitch name]
chhypervisor noderange [--removezfcpfrompool pool lun wwpn]
chhypervisor noderange [--smcli function arguments]
The chhypervisor command can be used to configure the RHEV-h.
The rhev-h host will register to the rhev-m automatically, but admin needs to approve the host can be added to the ‘cluster’ with -a flag .
After registering, the network interfaces of host need to be added to the ‘network’ of RHEV. And the power management for the host should be configured so that rhev-m could make proper decision when certain host encountered error.
The chhypervisor command can also be used to configure the zVM host.
For each host, an entry should be added to the hypervisor table:
The columns of hypervisor table:
hypervisor.node - rhev-h host name or zVM host name (lower-case).
hypervisor.type - Must be set to ‘rhevh’ or ‘zvm’.
hypervisor.mgr - The rhev manager (The FQDN of rhev-m server) for the host.
hypervisor.interface - The configuration for the nics. Refer to -n.
hypervisor.cluster - The cluster that the host will be added to. The default is ‘Default’ cluster if not specified.
-a Approve the host that to be added to cluster.
Before approve, the status of the host must be ‘pending_approval’.
-n Configure the network interfaces for the host.
Note: This operation only can be run when host is in ‘maintenance mode’. Use -d to switch the host to ‘maintenance’ mode.
The interfaces which configured in hypervisor.interface will be added to the network of RHEV.
The format of hypervisor.interface is multiple [network:interfacename: protocol:IP:netmask:gateway] sections separated with ‘|’. For example: [rhevm2:eth0:static:10.1.0.236:255.255.255.0:0.0.0.0].
network - The logic network which has been created by ‘cfgve -t nw’ or the default management network ‘rhevm’.
interfacename - Physical network name: ‘eth0’,’eth1’...
protocol - To identify which boot protocol to use for the interface: dhcp or static.
IP - The IP address for the interface.
netmask - The network mask for the interface.
gateway - The gateay for the interface. This field only can be set when the interface is added to ‘rhevm’ network.
-p Configure the power management for the host.
The power management must be configured for the rhev-h host to make the rhev-m to monitor the power status of the host, so that when certain host failed to function, rhev-m will fail over certain role like SPM to other active host.
For rack mounted server, the bmc IP and user:password need to be set for the power management (These parameters are gotten from ipmi table). rhev-m uses the ipmi protocol to get the power status of the host.
-e To activate the host.
-d To deactivate the host to maintenance mode.
-h Display usage message.
--adddisk2pool function region volume group
Add a disk to a disk pool defined in the EXTENT CONTROL. Function type can be either: (4) Define region as full volume and add to group OR (5) Add existing region to group. If the volume already exists in the EXTENT CONTROL, use function 5. If the volume does not exist in the EXTENT CONTROL, but is attached to SYSTEM, use function 4.
--addscsi device_number device_path option persist
Dynamically add a SCSI disk to a running z/VM system.
--addvlan name owner type transport
Create a virtual network LAN.
--addvswitch name osa_dev_addr osa_exp_adapter controller connect (0, 1, or 2) memory_queue router transport vlan_id port_type update gvrp native_vlan
Create a virtual switch.
--addzfcp2pool pool status wwpn lun size owner
Add a zFCP device to a device pool defined in xCAT. The device must have been carved up in the storage controller and configured with a WWPN/LUN before it can be added to the xCAT storage pool. z/VM does not have the ability to communicate directly with the storage controller to carve up disks dynamically.
--removediskfrompool function region group
Remove a disk from a disk pool defined in the EXTENT CONTROL. Function type can be either: (1) Remove region, (2) Remove region from group, (3) Remove region from all groups, OR (7) Remove entire group .
--removescsi device_number persist (YES or NO)
Delete a real SCSI disk.
--removevlan name owner
Delete a virtual network LAN.
--removevswitch name
Delete a virtual switch.
--removezfcpfrompool pool lun
Remove a zFCP device from a device pool defined in xCAT.
--smcli function arguments
Execute a SMAPI function. A list of APIs supported can be found by using the help flag, e.g. chhypervisor pokdev61 –smcli -h. Specific arguments associated with a SMAPI function can be found by using the help flag for the function, e.g. chhypervisor pokdev61 –smcli Image_Query_DM -h. Only z/VM 6.2 and older SMAPI functions are supported at this time. Additional SMAPI functions will be added in subsequent zHCP versions.
- To approve the host ‘host1’, enter:
chhypervisor host1 -a
- To configure the network interface for the host ‘host1’, enter:
chhypervisor host1 -n
- To configure the power management for the host ‘host1’, enter:
chhypervisor host1 -p
- To activate the host ‘host1’, enter:
chhypervisor host1 -e
- To deactivate the host ‘host1’, enter:
chhypervisor host1 -d
- To add a disk to a disk pool defined in the EXTENT CONTROL, enter:
chhypervisor pokdev61 --adddisk2pool 4 DM1234 DM1234 POOL1
- To add a zFCP device to a device pool defined in xCAT, enter:
chhypervisor pokdev61 --addzfcp2pool zfcp1 free 500501234567C890 4012345600000000 8G
- To remove a region from a group in the EXTENT CONTROL, enter:
chhypervisor pokdev61 --removediskfrompool 2 DM1234 POOL1
- To remove a zFCP device from a device pool defined in xCAT, enter:
chhypervisor pokdev61 --removezfcpfrompool zfcp1 4012345600000000 500501234567C890
- To execute a SMAPI function (Image_Query_DM), enter:
chhypervisor pokdev61 --smcli Image_Query_DM -T LNX3
/opt/xcat/bin/chhypervisor
chkkitcomp.1¶
chkkitcomp - Check if Kit components are compatible with an xCAT osimage.
chkkitcomp [-? | -h | --help] [-v | --version]
chkkitcomp [-V | --verbose] -i osimage kitcompname_list
The chkkitcomp command will check if the kit components are compatible with the xCAT osimage.
This command will ignore the current osimage.kitcomponents setting, and just to check if the kitcompname list in the cmdline are compatible with the osimage by osversion/ostype/osarch/ and kit component dependencies.
Note: The xCAT support for Kits is only available for Linux operating systems.
-h|--help
Display usage message.
-V|--verbose
Verbose mode.
-v|--version
Command version.
-i osimage
The name of the osimage is used for check.
kitcompname_list
A comma-delimited list of valid full kit component names or kit component basenames that are to be checking to the osimage.
- To check if a kit component is fitting to an osimage
chkkitcomp -i rhels6.2-ppc64-netboot-compute comp-test1-1.0-1-rhels-6.2-ppc64
Output is similar to:
Kit components comp-test1-1.0-1-rhels-6.2-ppc64 fit to osimage rhels6.2-ppc64-netboot-compute
lskit(1)|lskit.1, addkit(1)|addkit.1, rmkit(1)|rmkit.1, addkitcomp(1)|addkitcomp.1, rmkitcomp(1)|rmkitcomp.1
chkosimage.1¶
chkosimage - Use this xCAT command to check an xCAT osimage.
This command is currently supported for AIX osimages only.
Use this command to verify if the NIM lpp_source directories contain the correct software. The lpp_source directory must contain all the software that is specified in the “installp_bundle” and “otherpkgs” attributes of the osimage definition.
The command gets the name of the lpp_source resource from the xCAT osimage definition and the location of the lpp_source directory from the NIM resource definition.
It will check for installp, rpm and emgr type packages.
Note: Remember to use the prefixes, “I:”, “R:”, and “E:”, respectively, when specifying package names in an installp_bundle file or an otherpkgs list.
In addition to checking for missing software the chkosimage command will also check to see if there are multiple matches. This could happen when you use wildcards in the software file names. For example, if you have perl-xCAT* in a bundle file it could match multiple versions of the xCAT rpm package saved in your lpp_source directory.
If this happens you must remove the unwanted versions of the rpms. If the extra rpms are not removed you will get install errors.
To help with this process you can use the “-c|–clean” option. This option will keep the rpm package with the most recent timestamp and remove the others.
The chkosimage command should always be used to verify the lpp_source content before using the osimage to install any AIX cluster nodes.
-c |--clean
Remove any older versions of the rpms. Keep the version with the latest timestamp.
-h |--help
Display usage message.
osimage_name
The name of the xCAT for AIX osimage definition.
-V |--verbose
Verbose mode.
1
Check the XCAT osimage called “61image” to verify that the lpp_source directories contain all the software that is specified in the “installp_bundle” and “otherpkgs” attributes.
chkosimage -V 61image
2
Clean up the lpp_source directory for the osimage named “61img” by removing any older rpms with the same names but different versions.
chkosimage -c 61img
/opt/xcat/bin/chkosimage
This command is part of the xCAT software product.
mknimimage(1)|mknimimage.1
chvlan.1¶
chvlan - It adds or removes nodes for the vlan.
chvlan vlanid -n | --nodes noderange [-i | --interface nic]
chvlan vlanid -n | --nodes noderange -d | --delete
chvlan [-h | --help]
chvlan [-v | --version]
The chvlan command adds nodes to the given vlan. If -d is specified, the nodes will be removed from the vlan.
For added security, the root guard and bpdu guard will be enabled for the ports added to this vlan. However, the guards will not be disabled if the ports are removed from the vlan using chvlan (-d) or rmvlan commands. To disable them, you need to use the switch command line interface. Please refer to the switch command line interface manual to see how to disable the root guard and bpdu guard for a port.
vlanid is a unique vlan number.
-n|--nodes The nodes or groups to be added or removed. It can be stand alone nodes or KVM guests. It takes the noderange format. Please check the man page for noderange for details.
-i|--interface (For adding only). The interface name where the vlan will be tagged on. If omitted, the xCAT management network will be assumed. For KVM, it is the interface name on the host.
-h|--help Display usage message.
-v|--version The Command Version.
To add node1, node2 and node3 to vlan 3.
chvlan 3 -n node1,node2,node3
To add node1, node2 and node3 to vlan 3 using eth1 interface.
chvlan 3 -n node1,node2,node3 -i eth1
TO remove node1, node2 and node3 from vlan 3.
chvlan -n node1,node2,node3 -d
To add KVM guests node1 and node2 to vlan 3
mkdef node1 arch=x86_64 groups=kvm,all installnic=mac primarynic=mac mgt=kvm netboot=pxe nfsserver=10.1.0.204 os=rhels6 profile=compute provmethod=install serialport=0 serialspeed=115200 vmcpus=1 vmhost=x3650n01 vmmemory=512 vmnics=br0 vmstorage=nfs://10.1.0.203/vms mkdef node2 arch=x86_64 groups=kvm,all installnic=mac primarynic=mac mgt=kvm netboot=pxe nfsserver=10.1.0.204 os=rhels6 profile=compute provmethod=install serialport=0 serialspeed=115200 vmcpus=1 vmhost=x3650n01 vmmemory=512 vmnics=br0 vmstorage=nfs://10.1.0.203/vms chvlan 3 -n node1,node2 mkvm node1,node2 -s 20G rpower node1,node2 on rinstall node1,node2
To remove KVM guests node1 and node2 from vlan 3
chvlan 3 -n node1,node2 -d rpower node1,node2 off rmvm node1,node2
/opt/xcat/bin/chvlan
mkvlan(1)|mkvlan.1, rmvlan(1)|rmvlan.1, lsvlan(1)|lsvlan.1
chvlanports.1¶
chvlanports - It adds or removes nodes’ switch interfaces for the vlan.
chvlanports vlanid -n | --nodes noderange -i | --interface nic
chvlanports vlanid -n | --nodes noderange -i | --interface nic -d | --delete
chvlanports [-h | --help]
chvlanports [-v | --version]
The chvlanports command adds nodes switch interfaces to the given vlan. If -d is specified, the nodes switch interfaces will be removed from the vlan.
This command won’t create/remove vlans on switches, it just add node’s switch ports into exisitng vlan or remove them from existing vlan on switch. Before calling chvlanports, the nodes switch interfaces should be configured in table switch, and vlan must already existing in switches. =head1 Parameters
vlanid is a unique vlan number.
-n|--nodes The nodes or groups to be added or removed. It takes the noderange format. Please check the man page for noderange for details.
-i|--interface The interface name where the vlan will be tagged on.
-h|--help Display usage message.
-v|--version The Command Version.
To add node1, node2 and node3 to vlan 3 using eth1 interface.
chvlanports 3 -n node1,node2,node3 -i eth1
TO remove eth1 interface of node1, node2 and node3 from vlan 3.
chvlanports 3 -n node1,node2,node3 -i eth1 -d
/opt/xcat/bin/chvlanports
mkvlan(1)|mkvlan.1, rmvlan(1)|rmvlan.1, lsvlan(1)|lsvlan.1, chvlan(1)|chvlan.1
chvm.1¶
chvm - Changes HMC-, DFM-, IVM-, and zVM-managed partition profiles or virtual machines. For Power 775, chvm could be used to change the octant configuration values for generating LPARs; change the I/O slots assignment to LPARs within the same CEC.
chvm [-h| --help]
chvm [-v| --version]
chvm [-V| --verbose] noderange [-p profile]
chvm [-V| --verbose] noderange attr=val [attr=val...]
chvm noderange -*-p775* [**-p profile]
chvm noderange -*-p775* **-i id [-m memory_interleaving] -r partition_rule
chvm noderange [lparname={\* | name}]
- chvm noderange [vmcpus=min/req/max] [vmmemory=min/req/max]
- [vmothersetting=hugepage:N,bsr:N] [add_physlots=drc_index1,drc_index2...] [add_vmnics=vlan1[,vlan2..]] [add_vmstorage=<N|viosnode:slotid>] [--vios] [del_physlots=drc_index1,drc_index2...] [del_vadapter=slotid]
chvm noderange [-a size] [-d disk] [-p disk] [--resize disk=size] [--cpus count] [--mem memory]
chvm noderange [--add3390 disk_pool device_address size mode read_password write_password multi_password]
chvm noderange [--add3390active device_address mode]
chvm noderange [--add9336 disk_pool device_address size mode read_password write_password multi_password]
chvm noderange [--adddisk2pool function region volume group]
chvm noderange [--addnic device_address type device_count]
chvm noderange [--addpagespool volume_address volume_label volume_use system_config_name system_config_type parm_disk_owner parm_disk_number parm_disk_password]
chvm noderange [--addprocessor device_address]
chvm noderange [--addprocessoractive device_address type]
chvm noderange [--addvdisk device_address size]
chvm noderange [--addzfcp pool device_address loaddev size tag wwpn lun]
chvm noderange [--connectnic2guestlan device_address lan owner]
chvm noderange [--connectnic2vswitch device_address vswitch]
chvm noderange [--copydisk target_address source_node source_address]
chvm noderange [--dedicatedevice virtual_device real_device mode]
chvm noderange [--deleteipl]
chvm noderange [--disconnectnic device_address]
chvm noderange [--formatdisk device_address multi_password]
chvm noderange [--grantvswitch vswitch]
chvm noderange [--purgerdr]
chvm noderange [--removedisk device_address]
chvm noderange [--removenic device_address]
chvm noderange [--removeprocessor device_address]
chvm noderange [--removeloaddev wwpn lun]
chvm noderange [--removezfcp device_address wwpn lun]
chvm noderange [--replacevs directory_entry]
chvm noderange [--setipl ipl_target load_parms parms]
chvm noderange [--setpassword password]
chvm noderange [--setloaddev wwpn lun]
chvm noderange [--sharevolume volume_address share_enable]
chvm noderange [--undedicatedevice device_address]
The chvm command modifies the partition profile for the partitions specified in noderange. A partitions current profile can be read using lsvm, modified, and piped into the chvm command, or changed with the -p flag.
This command also supports to change specific partition attributes by specifying one or more “attribute equals value” pairs in command line directly, without whole partition profile.
For Power 755(use option –p775 to specify):
chvm could be used to change the octant configuration values for generating LPARs. chvm is designed to set the Octant configure value to split the CPU and memory for partitions, and set Octant Memory interleaving value. The chvm will only set the pending attributes value. After chvm, the CEC needs to be rebooted manually for the pending values to be enabled. Before reboot the cec, the administrator can use chvm to change the partition plan. If the the partition needs I/O slots, the administrator should use chvm to assign the I/O slots.
chvm is also designed to assign the I/O slots to the new LPAR. Both the current IO owning lpar and the new IO owning lpar must be powered off before an IO assignment. Otherwise, if the I/O slot is belonged to an Lpar and the LPAR is power on, the command will return an error when trying to assign that slot to a different lpar.
The administrator should use lsvm to get the profile content, and then edit the content, and add the node name with ”:” manually before the I/O which will be assigned to the node. And then the profile can be piped into the chvm command, or changed with the -p flag.
For normal power machine:
chvm could be used to modify the resources assigned to partitions. The admin shall specify the attributes with options vmcpus, vmmemory, add_physlots, vmothersetting, add_vmnics and/or add_vmstorage. If nothing specified, nothing will be returned.
The chvm command modifes the vm specified in noderange. Calling with deregister or purge options at the same time as the resize option is not recommended.
The chvm command modifes the virtual machine’s configuration specified in noderange.
-p profile
Name of an existing partition profile.
attr=val
Specifies one or more “attribute equals value” pairs, separated by spaces.
-V
Verbose output.
--p775
Specify the operation is for Power 775 machines.
-i
Starting numeric id of the newly created partitions. For Power 775 using Direct FSP Management, the id value only could be 1, 5, 9, 13, 17, 21, 25 and 29. Shall work with option --p775.
-m
memory interleaving. The setting value only could be 1 or 2. 2 means non-interleaved mode (also 2MC mode), the memory cannot be shared across the processors in an octant. 1 means interleaved mode (also 8MC mode) , the memory can be shared. The default value is 1. Shall work with option --p775.
-r
partition rule. Shall work with option --p775.
If all the octants configuration value are same in one CEC, it will be ” -r 0-7:value” .
If the octants use the different configuration value in one cec, it will be “-r 0:value1,1:value2,...7:value7”, or “-r 0:value1,1-7:value2” and so on.
The octants configuration value for one Octant could be 1, 2, 3, 4, 5. The meanings of the octants configuration value are as following:
1 -- 1 partition with all cpus and memory of the octant 2 -- 2 partitions with a 50/50 split of cpus and memory 3 -- 3 partitions with a 25/25/50 split of cpus and memory 4 -- 4 partitions with a 25/25/25/25 split of cpus and memory 5 -- 2 partitions with a 25/75 split of cpus and memory
-p profile
Name of I/O slots assignment profile. Shall work with option --p775.
lparname={\* | name}
Set LPAR name for the specified lpars. If ‘*’ specified, it means to get names from xCAT database and then set them for the specified lpars. If a string is specified, it only supports single node and the string will be set for the specified lpar. The user can use lsvm to check the lparnames for lpars.
vmcpus=value vmmemory=value add_physlots=value vmothersetting=value
To specify the parameters that will be modified.
add_vmnics=value add_vmstorage=value [--vios]
To create new virtual adapter for the specified node.
del_physlots=drc_index1,drc_index2...
To delete physical slots which are specified by the drc_index1,drc_index2....
del_vadapter=slotid
To delete a virtual adapter specified by the slotid.
-a size
Add a new Hard disk with size defaulting to GB. Multiple can be added with comma separated values.
--cpus count
Set the number of CPUs.
-d disk
Deregister the Hard disk but leave the backing files. Multiple can be done with comma separated values. The disks are specified by SCSI id. Size defaults to GB.
--mem memory
Set the memory, defaults to MB.
-p disk
Purge the Hard disk. Deregisters and deletes the files. Multiple can be done with comma separated values. The disks are specified by SCSI id. Size defaults to GB.
--resize disk=size
Change the size of the Hard disk. The disk can never be set to less than it’s current size. Multiple disks can be resized to size by using comma separated values on the left side of =. The disks are specified by SCSI id. Size defaults to GB.
--add3390 disk_pool device_address size mode read_password write_password multi_password
Adds a 3390 (ECKD) disk to a virtual machine’s directory entry. The device address can be automatically assigned by specifying ‘auto’. The size of the disk can be specified in GB, MB, or the number of cylinders.
--add3390active device_address mode
Adds a 3390 (ECKD) disk that is defined in a virtual machine’s directory entry to that virtual server’s active configuration.
--add9336 disk_pool device_address size mode read_password write_password multi_password
Adds a 9336 (FBA) disk to a virtual machine’s directory entry. The device address can be automatically assigned by specifying ‘auto’. The size of the disk can be specified in GB, MB, or the number of blocks.
--adddisk2pool function region volume group
Add a disk to a disk pool defined in the EXTENT CONTROL. Function type can be either: (4) Define region as full volume and add to group OR (5) Add existing region to group. The disk has to already be attached to SYSTEM.
--addnic device_address type device_count
Adds a network adapter to a virtual machine’s directory entry (case sensitive).
--addpagespool volume_addr volume_label volume_use system_config_name system_config_type parm_disk_owner parm_disk_number parm_disk_password
Add a full volume page or spool disk to the virtual machine.
--addprocessor device_address
Adds a virtual processor to a virtual machine’s directory entry.
--addprocessoractive device_address type
Adds a virtual processor to a virtual machine’s active configuration (case sensitive).
--addvdisk device_address size
Adds a v-disk to a virtual machine’s directory entry.
--addzfcp pool device_address loaddev size tag wwpn lun
Add a zFCP device to a device pool defined in xCAT. The device must have been carved up in the storage controller and configured with a WWPN/LUN before it can be added to the xCAT storage pool. z/VM does not have the ability to communicate directly with the storage controller to carve up disks dynamically. xCAT will find the a zFCP device in the specified pool that meets the size required, if the WWPN and LUN are not given. The device address can be automatically assigned by specifying ‘auto’. The WWPN/LUN can be set as the LOADDEV in the directory entry if (1) is specified as the ‘loaddev’.
--connectnic2guestlan device_address lan owner
Connects a given network adapter to a GuestLAN.
--connectnic2vswitch device_address vswitch
Connects a given network adapter to a VSwitch.
--copydisk target_address source_node source_address
Copy a disk attached to a given virtual server.
--dedicatedevice virtual_device real_device mode
Adds a dedicated device to a virtual machine’s directory entry.
--deleteipl
Deletes the IPL statement from the virtual machine’s directory entry.
--disconnectnic device_address
Disconnects a given network adapter.
--formatdisk disk_address multi_password
Formats a disk attached to a given virtual server (only ECKD disks supported). The disk should not be linked to any other virtual server. This command is best used after add3390().
--grantvswitch vswitch
Grant vSwitch access for given virtual machine.
--purgerdr
Purge the reader belonging to the virtual machine
--removedisk device_address
Removes a minidisk from a virtual machine’s directory entry.
--removenic device_address
Removes a network adapter from a virtual machine’s directory entry.
--removeprocessor device_address
Removes a processor from an active virtual machine’s configuration.
--removeloaddev wwpn lun
Removes the LOADDEV statement from a virtual machines’s directory entry.
--removezfcp device_address wwpn lun
Removes a given SCSI/FCP device belonging to the virtual machine.
--replacevs directory_entry
Replaces a virtual machine’s directory entry. The directory entry can be echoed into stdin or a text file.
--setipl ipl_target load_parms parms
Sets the IPL statement for a given virtual machine.
--setpassword password
Sets the password for a given virtual machine.
--setloaddev wwpn lun
Sets the LOADDEV statement in the virtual machine’s directory entry.
--undedicatedevice device_address
Delete a dedicated device from a virtual machine’s active configuration and directory entry.
- To change the partition profile for lpar4 using the configuration data in the file /tmp/lparfile, enter:
cat /tmp/lparfile | chvm lpar4
Output is similar to:
lpar4: Success
- To change the partition profile for lpar4 to the existing profile ‘prof1’, enter:
chvm lpar4 -p prof1
Output is similar to:
lpar4: Success
- To change partition attributes for lpar4 by specifying attribute value pairs in command line, enter:
chvm lpar4 max_mem=4096
Output is similar to:
lpar4: Success
- For Power 775, to create a new partition lpar1 on the first octant of the cec cec01, lpar1 will use all the cpu and memory of the octant 0, enter:
mkdef -t node -o lpar1 mgt=fsp groups=all parent=cec01 nodetype=lpar hcp=cec01
then:
chvm lpar1 --p775 -i 1 -m 1 -r 0:1
Output is similar to:
lpar1: Success
cec01: Please reboot the CEC cec1 firstly, and then use chvm to assign the I/O slots to the LPARs
- For Power 775, to create new partitions lpar1-lpar8 on the whole cec cec01, each LPAR will use all the cpu and memory of each octant, enter:
mkdef -t node -o lpar1-lpar8 nodetype=lpar mgt=fsp groups=all parent=cec01 hcp=cec01
then:
chvm lpar1-lpar8 --p775 -i 1 -m 1 -r 0-7:1
Output is similar to:
lpar1: Success
lpar2: Success
lpar3: Success
lpar4: Success
lpar5: Success
lpar6: Success
lpar7: Success
lpar8: Success
cec01: Please reboot the CEC cec1 firstly, and then use chvm to assign the I/O slots to the LPARs
- For Power 775 cec1, to create new partitions lpar1-lpar9, the lpar1 will use 25% CPU and 25% memory of the first octant, and lpar2 will use the left CPU and memory of the first octant. lpar3-lpar9 will use all the cpu and memory of each octant, enter:
mkdef -t node -o lpar1-lpar9 mgt=fsp groups=all parent=cec1 nodetype=lpar hcp=cec1
then:
chvm lpar1-lpar9 --p775 -i 1 -m 1 -r 0:5,1-7:1
Output is similar to:
lpar1: Success
lpar2: Success
lpar3: Success
lpar4: Success
lpar5: Success
lpar6: Success
lpar7: Success
lpar8: Success
lpar9: Success
cec1: Please reboot the CEC cec1 firstly, and then use chvm to assign the I/O slots to the LPARs
4.To change the I/O slot profile for lpar4 using the configuration data in the file /tmp/lparfile, the I/O slots information is similar to:
4: 514/U78A9.001.0123456-P1-C17/0x21010202/2/1
4: 513/U78A9.001.0123456-P1-C15/0x21010201/2/1
4: 512/U78A9.001.0123456-P1-C16/0x21010200/2/1
then run the command:
cat /tmp/lparfile | chvm lpar4 --p775
- To change the I/O slot profile for lpar1-lpar8 using the configuration data in the file /tmp/lparfile. Users can use the output of lsvm.and remove the cec information, and modify the lpar id before each I/O, and run the command as following:
chvm lpar1-lpar8 --p775 -p /tmp/lparfile
- To change the LPAR name, enter:
chvm lpar1 lparname=test_lpar01
Output is similar to:
lpar1: Success
- For Normal Power machine, to modify the resource assigned to a partition:
- Before modify, the resource assigned to node ‘lpar1’ can be shown with:
- lsvm lpar1
The output is similar to:
lpar1: Lpar Processor Info:
Curr Processor Min: 1.
Curr Processor Req: 4.
Curr Processor Max: 16.
lpar1: Lpar Memory Info:
Curr Memory Min: 1.00 GB(4 regions).
Curr Memory Req: 4.00 GB(16 regions).
Curr Memory Max: 32.00 GB(128 regions).
lpar1: 1,513,U78AA.001.WZSGVU7-P1-T7,0x21010201,0xc03(USB Controller)
lpar1: 1,512,U78AA.001.WZSGVU7-P1-T9,0x21010200,0x104(RAID Controller)
lpar1: 1/2/2
lpar1: 128.
To modify the resource assignment:
chvm lpar1 vmcpus=1/2/16 vmmemory=1G/8G/32G add_physlots=0x21010202
The output is similar to:
lpar1: Success
The resource information after modification is similar to:
lpar1: Lpar Processor Info:
Curr Processor Min: 1.
Curr Processor Req: 2.
Curr Processor Max: 16.
lpar1: Lpar Memory Info:
Curr Memory Min: 1.00 GB(4 regions).
Curr Memory Req: 8.00 GB(32 regions).
Curr Memory Max: 32.00 GB(128 regions).
lpar1: 1,514,U78AA.001.WZSGVU7-P1-C19,0x21010202,0xffff(Empty Slot)
lpar1: 1,513,U78AA.001.WZSGVU7-P1-T7,0x21010201,0xc03(USB Controller)
lpar1: 1,512,U78AA.001.WZSGVU7-P1-T9,0x21010200,0x104(RAID Controller)
lpar1: 1/2/2
lpar1: 128.
Note: The physical I/O resources specified with add_physlots will be appended to the specified partition. The physical I/O resources which are not specified but belonged to the partition will not be removed. For more information about add_physlots, please refer to lsvm(1)|lsvm.1.
chvm vm1 -a 8,16 --mem 512 --cpus 2
Output is similar to:
vm1: node successfully changed
- To adds a 3390 (ECKD) disk to a virtual machine’s directory entry:
chvm gpok3 --add3390 POOL1 0101 2G MR
Output is similar to:
gpok3: Adding disk 0101 to LNX3... Done
- To add a network adapter to a virtual machine’s directory entry:
chvm gpok3 --addnic 0600 QDIO 3
Output is similar to:
gpok3: Adding NIC 0900 to LNX3... Done
- To connects a given network adapter to a GuestLAN:
chvm gpok3 --connectnic2guestlan 0600 GLAN1 LN1OWNR
Output is similar to:
gpok3: Connecting NIC 0600 to GuestLan GLAN1 on LN1OWNR... Done
- To connects a given network adapter to a vSwitch:
chvm gpok3 --connectnic2vswitch 0600 VSW1
Output is similar to:
gpok3: Connecting NIC 0600 to vSwitch VSW1 on LNX3... Done
- To removes a minidisk from a virtual machine’s directory entry:
chvm gpok3 --removedisk 0101
Output is similar to:
gpok3: Removing disk 0101 on LNX3... Done
- To Removes a network adapter from a virtual machine’s directory entry:
chvm gpok3 --removenic 0700
Output is similar to:
gpok3: Removing NIC 0700 on LNX3... Done
- To replaces a virtual machine’s directory entry:
cat /tmp/dirEntry.txt | chvm gpok3 --replacevs
Output is similar to:
gpok3: Replacing user entry of LNX3... Done
/opt/xcat/bin/chvm
mkvm(1)|mkvm.1, lsvm(1)|lsvm.1, rmvm(1)|rmvm.1
chzone.1¶
chzone - Changes a zone defined in the cluster.
chzone <zonename> [--defaultzone] [-K] [-k full path to the ssh RSA private key] [-a noderange | -r noderange] [-g] [-f] [-s yes|no] [-V]
chzone [-h | -v]
The chzone command is designed to change the definition of a zone previous defined in the cluster. The chzone command is only supported on Linux ( No AIX support). The nodes are not updated with the new root ssh keys by chzone. You must run updatenode -k or xdsh -K to the nodes to update the root ssh keys to the new generated zone keys. This will also sync any service nodes with the zone keys, if you have a hierarchical cluster. Note: if any zones in the zone table, there must be one and only one defaultzone. Otherwise, errors will occur.
-h | --help
Displays usage information.
-v | --version
Displays command version and build date.
-k | --sshkeypath full path to the ssh RSA private key
This is the path to the id_rsa key that will be used to build new root’s ssh keys for the zone. If -k is used, it will generate the ssh public key from the input ssh RSA private key, and store both in /etc/xcat/sshkeys/<zonename>/.ssh directory.
-K | --genkeys
Using this flag, will generate new ssh RSA private and public keys for the zone into the /etc/xcat/sshkeys/<zonename>/.ssh directory. The nodes are not automatically updated with the new root ssh keys by chzone. You must run updatenode -k or xdsh -K to the nodes to update the root ssh keys to the new generated zone keys. This will also sync any service nodes with the zone keys, if you have a hierarchical cluster.
--default
if –defaultzone is input, then it will set the zone defaultzone attribute to yes. if –defaultzone is input and another zone is currently the default, then the -f flag must be used to force a change to the new defaultzone. If -f flag is not use an error will be returned and no change made. Note: if any zones in the zone table, there must be one and only one defaultzone. Otherwise, errors will occur.
-a | --addnoderange noderange
For each node in the noderange, it will set the zonename attribute for that node to the input zonename. If the -g flag is also on the command, then it will add the group name “zonename” to each node in the noderange.
-r | --rmnoderange noderange
For each node in the noderange, if the node is a member of the input zone, it will remove the zonename attribute for that node. If any of the nodes in the noderange is not a member of the zone, you will get an error and nothing will be changed. If the -g flag is also on the command, then it will remove the group name “zonename” from each node in the noderange.
-s| --sshbetweennodes yes|no
If -s entered, the zone sshbetweennodes attribute will be set to yes or no based on the input. When this is set to yes, then ssh will be setup to allow passwordless root access between nodes. If no, then root will be prompted for a password when running ssh between the nodes in the zone.
-f | --force
Used with the (–defaultzone) flag to override the current default zone.
-g | --assigngroup
Used with the (-a or -r ) flag to add or remove the group zonename for all nodes in the input noderange.
-V | --Verbose
Verbose mode.
*
*
To generate new root ssh keys for zone2A using the ssh id_rsa private key in /root/.ssh:
chzone zone2A -k /root/.ssh
Note: you must use xdsh -K or updatenode -k to update the nodes with the new keys
*
To generate new root ssh keys for zone2A, enter :
chzone zone2A -K
Note: you must use xdsh -K or updatenode -k to update the nodes with the new keys
*
To add a new group of nodes (compute3) to zone3 and add zone3 group to the nodes, enter:
chzone zone3 -a compute3 -g
*
To remove a group of nodes (compute4) from zone4 and remove zone4 group from the nodes, enter:
chzone zone4 -r compute4 -g
*
To change the sshbetweennodes setting on the zone to not allow passwordless ssh between nodes, enter:
chzone zone5 -s no
Note: you must use xdsh -K or updatenode -k to update the nodes with this new setting.
Files
/opt/xcat/bin/chzone/
Location of the chzone command.
L <mkzone(1)|mkzone.1>,L <rmzone(1)|rmzone.1>,L <xdsh(1)|xdsh.1>, updatenode(1)|updatenode.1
clonevm.1¶
clonevm - Create masters from virtual machines and virtual machines from masters.
clonevm noderange [ -t <mastertobemade | -b <master to base vms upon> ] -d|–detached -f|–force>
Command to promote a VM’s current configuration and storage to a master as well as performing the converse operation of creating VMs based on a master.
By default, attempting to create a master from a running VM will produce an error. The force argument will request that a master be made of the VM anyway.
Also, by default a VM that is used to create a master will be rebased as a thin clone of that master. If the force argument is used to create a master of a powered on vm, this will not be done. Additionally, the detached option can be used to explicitly request that a clone not be tethered to a master image, allowing the clones to not be tied to the health of a master, at the cost of additional storage.
When promoting a VM’s current state to master, all rleated virtual disks will be copied and merged with any prerequisite images. A master will not be tethered to other masters.
-h|--help Display usage message.
-b The master to base the clones upon
-t The target master to copy a single VM’s state to
-d Explicitly request that the noderange be untethered from any masters.
-f Force cloning of a powered on VM. Implies -d if the VM is on.
-v|--version Command Version.
-V|--verbose Verbose output.
Creating a master named appserver from a node called vm1: clonevm vm1 -t appserver
Cleating 30 VMs from a master named appserver: clonevm vm1-vm30 -b appserver
/opt/xcat/bin/clonevm
chvm(1)|chvm.1, lsvm(1)|lsvm.1, rmvm(1)|rmvm.1, mkvm(1)|mkvm.1, vmmaster(5)|vmmaster.5
configfpc.1¶
configfpc - discover the Fan Power Controllers (FPCs) and configure the FPC interface
configfpc -i interface
configfpc -i interface --ip default ip address
configfpc [-V | --verbose]
configfpc [-h | --help | -?]
configfpc will discover and configure all FPCs that are set to the default IP address. If not supplied the default ip is 192.168.0.100.
The -i interface is required to direct configfpc to the xCAT MN interface which is on the same VLAN as the FPCs.
There are several bits of information that must be included in the xCAT database before running this command.
You must create the FPC node definitions for all FPCs being discovered including the IP address and switch port information.
The configfpc command discovers the FPCs and collects the MAC address. The MAC address is used to relate the FPC to a FPC node using the switch information for this MAC. Once the relationship is discovered the FPC is configured with the FPC node IP settings.
This process is repeated until no more FPCs are discovered.
For more information on xCAT support of NeXtScale and configfpc see the following doc: XCAT_NeXtScale_Clusters
-i interface
Use this flag to specify which xCAT MN interface (example: eth4) that is connected to the NeXtScale FPCs. This option is required.
--ip default ip address
Use this flag to override the default ip address of 192.168.0.100 with a new address.
-V | --verbose
Verbose mode
1
To discover and configure all NeXtScale Fan Power Controllers (FPCs) connected on eth0 interface.
configfpc -i eth0
2
To override the default ip address and run in Verbose mode.
configfpc -i eth0 --ip 196.68.0.100 -V
csm2xcat.1¶
csm2xcat - Allows the migration of a CSM database to an xCAT database.
The csm2xcat command must be run on the Management Server of the CSM system that you want to migrate to xCAT. The commmand will build two xCAT stanza files that can update the xCAT database with the chdef command.
Copy the csm2xcat command to the CSM Management Server. Run the command, indicating where you want your stanza files saved with the –dir parameter. Check the stanza files to see if the information is what you want put in the xCAT database. Copy the two stanza files: node.stanza, device.stanza back to your xCAT Management node, and run the chdef command to input into the xCAT database.
- To build xCAT stanza files, enter on the CSM Management Server:
csm2xcat --dir /tmp/mydir
- To put the data in the xCAT database on the xCAT Management Node:
cat node.stanza | chdef -z
cat device.stanza | chdef -z
chdef(1)|chdef.1
db2sqlsetup.1¶
db2sqlsetup - Sets up the IBM DB2 for xCAT to use.
db2sqlsetup {-h | --help}
db2sqlsetup {-v | --version}
db2sqlsetup {-i | --init}{<-S> | <-C>} [-o|-**-setupODBC] [**-V | --verbose]
db2sqlsetup {-i | --init}{<-S>} [-N|-**-nostart] [-o|--setupODBC] [**-V | --verbose]
db2sqlsetup {-o | --setupODBC} {<-S> | <-C>} [-V|-****-verbose]
db2sqlsetup {-p | --passwd} [<-S> | <-C>]
db2sqlsetup - Sets up the IBM DB2 database for xCAT to use. The db2sqlsetup script is run on the Management Node, after the DB2 Server code has been installed, to setup the DB2 Server (-S). The xcatd daemon will be stopped during migration on the MN. No xCAT commands should be run during the init process, because we will be migrating the xCAT database to DB2 and restarting the xcatd daemon.
The db2sqlsetup script must be run on each Service Node, after the DB2 Client code has been installed, to setup the DB2 Client (-C). There are two postscripts that are provided ( db2install and odbcsetup) that will automatically setup you Service Node as a DB2 client.
For full information on the setup of DB2, see Setting_Up_DB2_as_the_xCAT_DB.
When running of db2sqlsetup on the MN: One password must be supplied for the setup, a password for the xcatdb unix id which will be used as the DB2 instance id and database name. The password will be prompted for interactively or can be input with the XCATDB2PW environment variable. The script will create the xcat database instance (xcatdb) in the /var/lib/db2 directory unless overriden by setting the site.databaseloc attribute. This attribute should not be set to the directory that is defined in the installloc attribute and it is recommended that the databaseloc be a new filesystem dedicated to the DB2 database, especially in very large clusters.
When running db2sqlseutp on the SN: Not only will the password for the DB2 instance Id be prompted for and must match the one on the Management Node; but also the hostname or ip address of the Management Node as known by the Service Node must be supplied , unless the XCATDB2SERVER environment variable is set. You can automatically install and setup of DB2 on the SN using the db2install and odbcsetup postscripts and not need to manually run the command. See the full documentation.
Note: On AIX , root must be running ksh and on Linux, bash shell.
-h|--help
Displays the usage message.
-v|--version
Displays the release version of the code.
-V|--verbose
Displays verbose messages.
-i|--init
The init option is used to setup an installed DB2 database on AIX or Linux (p-Series) so that xCAT can use the database. This must be combined with either the -S or -C flag to indicate whether we are setting up the Server or the Client. With the -S flag, it involves creating the xcatdb database, the xcatdb instance id, allowing access to the xcatdb database by the Management Node. It also backs up the current xCAT database and restores it into the newly setup xcatdb DB2 database. It creates the /etc/xcat/cfgloc file to point the xcatd daemon to the DB2 database and restarts the xcatd daemon using the database.
-p|--passwd
The password change option is to change the database access password for the DB2 xcatdb database. If -S is input then it will only change the password on the DB2 Server (MN). If -C is input it will only change on the DB2 clients (SN). If neither -S or -C are input with this flag, then it will change both the DB2 Server and Clients. When changing the password the xcatd daemon will be stopped and restarted. Any other tools accessing the database should also be stopped before changing and restarted after changing.
-S|-C
This options says whether to setup the Server (-S) on the Management Node, or the Client (-C) on the Service Nodes.
-N|--nostart
This option with the -S flag will create the database, but will not backup and restore xCAT tables into the database. It will create the cfgloc file such that the next start of xcatd will try and contact the database. This can be used to setup the xCAT DB2 database during or before install.
-o|--setupODBC
This option sets up the ODBC /etc/../odbcinst.ini, /etc/../odbc.ini and the .odbc.ini file in roots home directory will be created and initialized to run off the xcatdb DB2 database.
*
XCATDB2INSPATH overrides the default install path for DB2 which is /opt/ibm/db2/V9.7 for Linux and /opt/IBM/db2/V9.7 for AIX.
*
DATABASELOC override the where to create the xcat DB2 database, which is /var/lib/db2 by default of taken from the site.databaseloc attribute.
*
XCATDB2PW can be set to the password for the xcatdb DB2 instance id so that there will be no prompting for a password when the script is run.
*
To setup DB2 Server for xCAT to run on the DB2 xcatdb database, on the MN:
db2sqlsetup -i -S
*
To setup DB2 Client for xCAT to run on the DB2 xcatdb database, on the SN:
db2sqlsetup -i -C
*
To setup the ODBC for DB2 xcatdb database access, on the MN :
db2sqlsetup -o -S
*
To setup the ODBC for DB2 xcatdb database access, on the SN :
db2sqlsetup -o -C
*
To setup the DB2 database but not start xcat running with it:
db2sqlsetup -i -S -N
*
To change the DB2 xcatdb password on both the Management and Service Nodes:
db2sqlsetup -p
dumpxCATdb.1¶
dumpxCATdb - dumps the xCAT db tables .
dumpxCATdb [-a] [-V] [{-p | --path} path]
dumpxCATdb [-b] [-V] [{-p | --path} path]
dumpxCATdb [-h | --help] [-v | --version]
If not using the binary dump option (-b), then the dumpxCATdb command creates .csv files for xCAT database tables and puts them in the directory given by the -p flag. These files can be used by the restorexCATdb command to restore the database. The command will read the list of tables in the site.skiptables attribute and not backup those tables. Supports using XCAT_SKIPTABLES env variable to provide a list of skip tables. The command will never backup TEAL or ISNM tables, except isnm_config. To dump TEAL tables use the documented process for TEAL. For ISNM use tabdump, after using tabprune to get to prune unnecessary records.
If using the binary dump option for the DB2 or postgreSQL database, then the routine will use the Database provide utilites for backup of the entire database.
-h Display usage message.
-v Command Version.
-V Verbose.
-a All,without this flag the eventlog and auditlog will be skipped.
-b This flag is only used for the DB2 or postgreSQL database. The routine will use the database backup utilities to create a binary backup of the entire database. Note to use this backup on DB2, you will have first had to modify the logging of the database and have taken an offline initial backup. Refer to the xCAT DB2 documentation for more instructions.
-p Path to the directory to dump the database. It will be created, if it does not exist.
- To dump the xCAT database into the /tmp/db directory, enter:
dumpxCATdb -p /tmp/db
- To dump the xCAT database into the /tmp/db directory, including the auditlog and eventlog enter:
dumpxCATdb -a -p /tmp/db
- To have dumpxCATdb not backup the hosts or passwd table:
chtab key=skiptables site.value=”hosts,passwd”
dumpxCATdb -p /tmp/db
- To have dumpxCATdb not backup the hosts or passwd table:
export XCAT_SKIPTABLES=”hosts,passwd”
dumpxCATdb -p /tmp/db
- To have dumpxCATdb use DB2 utilities to backup the DB2 database:
dumpxCATdb -b -p /install/db2backup
/opt/xcat/sbin/dumpxCATdb
restorexCATdb(1)|restorexCATdb.1
genimage.1¶
genimage - Generates a stateless image to be used for a diskless install.
genimage
genimage [-o osver] [-a arch] [-p profile] [-i nodebootif] [-n nodenetdrivers] [--onlyinitrd] [-r otherifaces] [-k kernelver] [-g krpmver] [-m statelite] [-l rootlimitsize] [--permission permission] [--interactive] [--dryrun] [--ignorekernelchk] [--noupdate] imagename
genimage -o osver [-a arch] -p profile -i nodebootif -n nodenetdrivers [--onlyinitrd] [-r otherifaces] [-k kernelver] [-g krpmver] [-m statelite] [-l rootlimitsize] [--permission permission] [--interactive] [--dryrun] [--noupdate]
genimage [-h | --help | -v | --version]
Generates a stateless and a statelite image that can be used to boot xCAT nodes in a diskless mode.
The imagename format of the command is recommended. When specified, genimage will use the osimage definition for information to generate this image. Additional options specified on the command line will override any corresponding previous osimage settings, and will be written back to the osimage definition.
- If imagename is not specified (old method):
- the default packages included (and excluded) in the image are specified by
/opt/xcat/share/xcat/netboot/<os>/<profile>[.<osver>][.<arch>].pkglist and
/opt/xcat/share/xcat/netboot/<os>/<profile>[.<osver>][.<arch>].exlist.
- Additional packages that are not from the os distro can be specified in a
/opt/xcat/share/xcat/netboot/<os>/<profile>[.<osver>][.<arch>].otherpkgs.pkglist file.
- Customized package list files will override these files and can be specified under /install/custom/netboot/<os> directory.
- The generated image will be put in /install/netboot/<osver>/<arch>/<profile> directory.
- osimage definitions will be created in the I<linuximage> and I<osimage> tables. The newly generated image names will have the following format:
for stateless: <osver>-<arch>-netboot-<profile>
for statelite: <osver>-<arch>-statelite-<profile>
If genimage runs on the management node, both the osimage table and linuximage table will be updated with the given values from the options.
The genimage command will generate two initial ramdisks for stateless and statelite, one is initrd-stateless.gz, the other one is initrd-statelite.gz.
After your image is generated, you can chroot to the image, install any additional software you would like, or make modifications to files, and then run the following command to prepare the image for deployment.
for stateless: packimage
for statelite: liteimg
Besides prompting for some paramter values, the genimage command takes default guesses for the parameters not specified or not defined in the osimage and linuximage tables. It also assumes default answers for questions from the yum/zypper command when installing rpms into the image. Please use -****-interactive flag if you want the yum/zypper command to prompt you for the answers.
If --onlyinitrd is specified, genimage only regenerates the initrd for a stateless image to be used for a diskless install.
The genimage command must be run on a system that is the same architecture and same distro with same major release version as the nodes it will be used on. If the management node is not the same architecture or same distro level, copy the contents of /opt/xcat/share/xcat/netboot/<os> to a system that is the proper architecture, and mount /install from the management node to that system. Then change directory to /opt/xcat/share/xcat/netboot/<os> and run ./genimage.
imagename specifies the name of an os image definition to be used. The specification for the image is stored in the osimage table and linuximage table.
-a arch
The hardware architecture of this node: x86_64, ppc64, x86, ia64, etc. If omitted, the current hardware architecture will be used.
-o osver
The operating system for the image: fedora8, rhel5, sles10, etc. The OS packages must be in /install/<osver>/<arch> (use copycds(8)|copycds.8).
-p profile
The profile (e.g. compute, service) to use to create the image. This determines what package lists are used from /opt/xcat/share/xcat/netboot/<os> to create the image with. When deploying nodes with this image, the nodes’ nodetype.profile attribute must be set to this same value.
-i nodebootif
This argument is now optional, and allows you to specify the network boot interface to be configured in the image (e.g. eth0). If not specified, the interface will be determined and configured during the network boot process.
-n nodenetdrivers
This argument is now optional, and allows you to specify the driver modules needed for the network interface(s) on your stateless nodes. If you do not specify this option, the default is to include all recent IBM xSeries network drivers.
If specified, nodenetdrivers should be a comma separated list of network drivers to be used by the stateless nodes (Ie.: -n tg3,e1000). Note that the drivers will be loaded in the order that you list them, which may prove important in some cases.
-l rootlimit
The maximum size allowed for the root file system in the image. Specify in bytes, or can append k, m, or g.
--onlyinitrd
Regenerates the initrd for a stateless image to be used for a diskless install.
Regenerates the initrd that is part of a stateless/statelite image that is used to boot xCAT nodes in a stateless/stateli te mode. The generated initrd will be put in /install/netboot/<OS>/<arch>/<profile>.
The genimage --onlyinitrd command will generate two initial ramdisks, one is initrd-statelite.gz for statelite mode, the other one is initrd-stateless.gz for stateless mode.
--permission permission
The mount permission of /.statelite directory for statelite mode, which is only used for statelite mode, and the default permission is 755.
-r otherifaces
Other network interfaces (e.g. eth1) in the image that should be configured via DHCP.
-k kernelver
Use this flag if you want to use a specific version of the kernel in the image. Defaults to the first kernel found in the install image.
-g krpmver
Use this flag to specify the rpm version for kernel packages in the image. It must be present if -k flag is specified in the command for SLES. Generally, the value of -g is the part after linux- and before .rpm in a kernel rpm name.
-m statelite
This flag is for Ubuntu, Debian and Fedora12 only. Use this flag to specify if you want to generate statelite image. The default is to generate stateless image for these three operating systems. For others, this flag is invalid because both stateless and statelite images will be generated with this command.
--interactive
This flag allows the user to answer questions from yum/zypper command when installing rpms into the image. If it is not specified, ‘-y’ will be passed to the yum command and ‘–non-interactive –no-gpg-checks’ will be passed to the zypper command as default answers.
--dryrun
This flag shows the underlying call to the os specific genimage function. The user can copy and the paste the output to run the command on another machine that does not have xCAT installed.
-t tmplimit
(Deprecated) This flag allows the user to setup the /tmp and the /var/tmp file system sizes. This flag is no longer supported. You can overwrite any file system size using the .postinstall script where you can create a new /etc/fstab file.
--ignorekernelchk
Skip the kernel version checking when injecting drivers from osimage.driverupdatesrc. That means all drivers from osimage.driverupdatesrc will be injected to initrd for the specific target kernel.
--noupdate
This flag allows the user to bypass automatic package updating when installing other packages.
-v|--version
Display version.
-h|--help
Display usage message.
1 To prompt the user for inputs:
genimage
2 To generate an image using information from an osimage definition:
genimage myimagename
3 To run genimage in test mode without actually generating an image:
genimage --dryrun myimagename
4 To generate an image and have yum/zypper prompt for responses:
genimage myimagename --interactive
5 To generate an image, replacing some values in the osimage definition:
genimage -i eth0 -n tg3 myimagename
6 (old method) To generate a fedora8 image for a compute node architecture x86_64 and place it in the /install/netboot/fedora8/x86_64/compute/rootimg directory:
genimage -i eth0 -o fedora8 -p compute
7 (old method)
genimage -i eth0 -r eth1,eth2 -n tg3,bnx2 -o centos5.1 -p compute
8 (old method)
genimage -i eth0 -n tg3,bnx2 -o sles11 -p compute --interactive
9 (old method)
genimage -i eth0 -n igb,e1000e,e1000,bnx2,tg3 -o centos5.4 -p nfsroot --permission 777
10 (old method) To regenerate the initrd for a fedora8 image for a compute node architecture x86_64 and place it in the /install/netboot/fedora8/x86_64/compute/rootimg directory:
cd /opt/xcat/share/xcat/netboot/fedora ./genimage --onlyinitrd -i eth0 -n tg3,bnx2 -o fedora8 -p compute
packimage(1)|packimage.1, liteimg(1)|liteimg.1
geninitrd.1¶
geninitrd - Generate an initrd (initial ramfs) which to be used for statefull install or stateless netboot.
Generate the initrd for the osimage: imagename which is an xCAT object of osimage type.
Diskfull Osimage
If the imagename is a statefull one (The provmethod attribute for the osimage is ‘install’), this command is used to rebuild the initrd to inject the new drivers from driver rpms or ‘update distro’ and copy the rebuilt initrd and new kernel (If there’s new kernel in ‘update distro’) to the directory /tftpboot/xcat/<imagename>.
If the initrd has been rebuilt by geninitrd, when run nodeset, the –noupdateinitrd option should be used to skip the rebuilding of initrd to improve the performance.
Three attributes of osimage object can be used to specify the Driver RPM location and Driver names for injecting new drviers to initrd.
netdrivers - comma separated driver names that need to be injected to the initrd. The postfix ‘.ko’ can be ignored. The netdrivers attribute must be set to specify the new driver list. If you want to load all the drivers from the driver rpms, using the keyword allupdate.
driverupdatesrc - comma separated driver rpm packages (full path should be specified)
osupdatename - comma separated ‘osdistroupdate’ object. Each ‘osdistroupdate’ object specifies a Linux distro update. When run geninitrd, ‘kernel-*.rpm’ will be searched from osdistroupdate.dirpath to get all the rpm packages and then search the drivers from the rpm packages.
Refer to the doc: Using_Linux_Driver_Update_Disk
Stateless Osimage
If the imagename is a stateless one (The provmethod attribute for the osimage is ‘netboot’), this command is used to generate the initrd from the rootimg which generated by ‘genimage’ command. So the ‘genimage’ must be run once before running the geninitrd command.
Two attributes of osimage object can be used to specify the Driver RPM location and Driver names for injecting new drviers to initrd.
netdrivers - comma separated driver names that need to be injected to the initrd. The postfix ‘.ko’ can be ignored. The netdrivers attribute must be set to specify the new driver list. If you want to load all the drivers from the driver rpms, using the keyword allupdate.
driverupdatesrc - comma separated driver rpm packages (full path should be specified)
imagename specifies the name of an os image definition to be used. The specification for the image is storted in the osimage table and linuximage table.
--ignorekernelchk
Skip the kernel version checking when injecting drivers from osimage.driverupdatesrc. That means all drivers from osimage.driverupdatesrc will be injected to initrd for the specific target kernel.
geninitrd(1)|geninitrd.1, genimage(1)|genimage.1
getmacs.1¶
getmacs - Collects node MAC address.
getmacs [-h| --help | -v| --version]
getmacs noderange [-F filter]
getmacs noderange [-M]
getmacs noderange [-V| --verbose] [-f] [-d] [--arp] | [-D [-S server] [-G gateway] [-C client] [-o]]
The getmacs command collects MAC address from a single or range of nodes. Note that on AIX systems, the returned MAC address is not colon-seperated (for example 8ee2245cf004), while on Linux systems the MAC address is colon-seperated (for example 8e:e2:24:5c:f0:04). If no ping test performed, getmacs writes the first adapter MAC to the xCAT database. If ping test performed, getmacs will write the first successfully pinged MAC to xCAT database.
For PPC (using Direct FSP Management) specific:
Note: If network adapters are physically assigned to LPARs, getmacs cannot read the MAC addresses unless perform ping test, since there is no HMC command to read them and getmacs has to login to open formware. And if the LPARs has never been activated before, getmacs need to be performed with the option “-D” to get theirs MAC addresses.
For PPC (using HMC) specific:
Note: The option “-D” must be used to get MAC addresses of LPARs.
For IBM Flex Compute Node (Compute Node for short) specific:
Note: If “-d” is specified, all the MAC of the blades will be displayed. If no option specified, the first MAC address of the blade will be written to mac table.
--arp
Read MAC address with ARP protocal.
-C
Specify the IP address of the partition for ping test. The default is to read from xCAT database if no -C specified.
-d
Display MAC only. The default is to write the first valid adapter MAC to the xCAT database.
-D
Perform ping test. Please be aware that in this way, the lpars will be reset.
-f
Force immediate shutdown of the partition.This flag must be used with -D flag.
-F
Specify filters to select the correct adapter. Acceptable filters are Type, MAC_Address, Phys_Port_Loc, Adapter, Port_Group, Phys_Port, Logical_Port, VLan, VSwitch, Curr_Conn_Speed.
-G
Gateway IP address of the partition. The default is to read from xCAT database if no -G specified.
-h
Display usage message.
-M
Return multiple MAC addresses for the same adapter or port, if available from the hardware. For some network adapters (e.g. HFI) the MAC can change when there are some recoverable internal errors. In this case, the hardware can return several MACs that the adapter can potentially have, so that xCAT can put all of them in DHCP. This allows successful booting, even after a MAC change, but on Linux at this time, it can also cause duplicate IP addresses, so it is currently not recommended on Linux. By default (without this flag), only a single MAC address is returned for each adapter.
-o
Read MAC address when the lpar is in openfirmware state. This option mush be used with [-D] option to perform ping test. Before use -o, the lpar must be in openfirmware state.
-S
The IP address of the machine to ping. The default is to read from xCAT databse if no -S specified.
-v
Command Version.
-V
Verbose output.
-i
Specify the interface whose mac address will be collected and written into mac table. If 4 mac addresses are returned by option ‘-d’, they all are the mac addresses of the blade. The N can start from 0(map to the eth0 of the blade) to 3. If 5 mac addresses are returned, the 1st mac address must be the mac address of the blade’s FSP, so the N will start from 1(map to the eth0 of the blade) to 4.
0 The command completed successfully.
1 An error has occurred.
- To retrieve the MAC address for the HMC-managed partition lpar4 and write the first valid adapter MAC to the xCAT database, enter:
getmacs lpar4
Output is similar to:
lpar4:
#Type MAC_Address Phys_Port_Loc Adapter Port_Group Phys_Port Logical_Port VLan VSwitch Curr_Conn_Speed
hea 7607DFB07F02 N/A N/A N/A N/A N/A 1 ETHERNET0 N/A
ent U78A1.001.99203B5-P1-T6 00145eb55788 /lhea@23c00614/ethernet@23e00514 unsuccessful physical
- To retrieve the MAC address with ARP protocal:
getmacs lpar4 --arp
Output is similar to:
lpar4:
#IP MAC_Address
192.168.0.10 00145eb55788
- To retrieve the MAC address for the HMC-managed partition lpar4 and display the result only, enter:
getmacs lpar4 -d
Output is similar to:
lpar4:
#Type MAC_Address Phys_Port_Loc Adapter Port_Group Phys_Port Logical_Port VLan VSwitch Curr_Conn_Speed
hea 7607DFB07F02 N/A N/A N/A N/A N/A 1 ETHERNET0 N/A
ent U78A1.001.99203B5-P1-T6 00145eb55788 /lhea@23c00614/ethernet@23e00514 unsuccessful physical
- To retrieve the MAC address for the HMC-managed partition lpar4 with filter Type=hea,VSwitch=ETHERNET0.
getmacs lpar4 -F Type=hea,VSwitch=ETHERNET0
Output is similar to:
lpar4:
#Type MAC_Address Phys_Port_Loc Adapter Port_Group Phys_Port Logical_Port VLan VSwitch Curr_Conn_Speed
hea 7607DFB07F02 N/A N/A N/A N/A N/A 1 ETHERNET0 N/A
- To retrieve the MAC address while performing a ping test for the HMC-managed partition lpar4 and display the result only, enter:
getmacs lpar4 -d -D -S 9.3.6.49 -G 9.3.6.1 -C 9.3.6.234
Output is similar to:
lpar4:
#Type Location Code MAC Address Full Path Name Ping Result
ent U9133.55A.10B7D1G-V12-C4-T1 8e:e2:24:5c:f0:04 /vdevice/l-lan@30000004 successful virtual
- To retrieve the MAC address for Power 775 LPAR using Direct FSP Management without ping test and display the result only, enter:
getmacs lpar4 -d
Output is similar to:
lpar4:
#Type Phys_Port_Loc MAC_Address Adapter Port_Group Phys_Port Logical_Port VLan VSwitch Curr_Conn_Speed
HFI N/A 02:00:02:00:00:04 N/A N/A N/A N/A N/A N/A N/A
- To retrieve multiple MAC addresses from Power 775 HFI network adapter using Direct FSP Management, enter:
getmacs lpar4 -M
Output is similar to:
lpar4:
#Type Phys_Port_Loc MAC_Address Adapter Port_Group Phys_Port Logical_Port VLan VSwitch Curr_Conn_Speed
HFI N/A 02:00:02:00:00:04|02:00:02:00:00:05|02:00:02:00:00:06 N/A N/A N/A N/A N/A N/A N/A
/opt/xcat/bin/getmacs
makedhcp(8)|makedhcp.8
getslnodes.1¶
getslnodes - queries your SoftLayer account and gets attributes for each server.
The getslnodes command queries your SoftLayer account and gets attributes for each server. The attributes can be piped to ‘mkdef -z’ to define the nodes in the xCAT DB so that xCAT can manage them.
Before using this command, you must download and install the SoftLayer API perl module. For example:
cd /usr/local/lib
git clone https://github.com/softlayer/softlayer-api-perl-client.git
You also need to follow these directions to get your SoftLayer API key: http://knowledgelayer.softlayer.com/procedure/retrieve-your-api-key
getslnodes requires a .slconfig file in your home directory that contains your SoftLayer userid, API key, and location of the SoftLayer API perl module, in attr=val format. For example:
# Config file used by the xcat cmd getslnodes
userid = joe_smith
apikey = 1234567890abcdef1234567890abcdef1234567890abcdef
apidir = /usr/local/lib/softlayer-api-perl-client
Display information about all of the nodes in your SoftLayer account:
getslnodesDisplay information about all of the nodes whose hostname starts with foo:
getslnodes foo
Create xCAT node defintions in the xCAT DB for all of the nodes in your SoftLayer account:
getslnodes | mkdef -z
/opt/xcat/bin/getslnodes
pushinitrd(1)|pushinitrd.1
gettab.1¶
gettab - select table rows, based on attribute criteria, and display specific attributes.
gettab [-H | --with-fieldname] key=value,... table.attribute ...
gettab [-? | -h | --help]
The gettab command uses the specified key values to select a row in each of the tables requested. For each selected row, the specified attributes are displayed. The gettab command can be used instead of nodels for tables that are not keyed by nodename (e.g. the site table), or to select rows based on an attribute value other than nodename.
-H|--with-fieldname
Always display table.attribute name next to result. By default, this is done only if more than one table.attribute is requested.
-?|-h|--help
Display usage message.
*
To display setting for master (management node) in the site table:
gettab -H key=master site.value
The output would be similar to:
site.value: mgmtnode.cluster.com
*
To display the first node or group name that has mgt set to blade in the nodehm table:
gettab mgt=blade nodehm.node
The output would be similar to:
blades
/opt/xcat/bin/gettab
nodels(1)|nodels.1, chtab(8)|chtab.8, tabdump(8)|tabdump.8
getxcatdocs.1¶
getxcatdocs - downloads the xCAT documentation and converts to HTML and PDF
getxcatdocs [-? | -h | --help] getxcatdocs [-v | --verbose] [destination-dir] getxcatdocs [-v | --verbose] [-c | --continue] [-d | --doc single_doc] [destination-dir]
The getxcatdocs command downloads the xCAT documentation from the wiki and converts it to both HTML and PDF. This enables reading the documentation when you do not have internet access. Note that this command does not download/convert the entire xCAT wiki - only the “official” xCAT documentation linked from http://sourceforge.net/p/xcat/wiki/XCAT_Documentation.
If destination-dir is specified, getxcatdocs will put the converted documentation in that directory, in 3 sub-directories: html, pdf, images. Otherwise, it will put it in the current directory (in the same three sub-directories).
If --doc single_doc is specified, only that one wiki page will be downloaded and converted.
getxcatdocs uses curl to run the Allura wiki API to download the document markdown text, and Pandoc with LaTex them to PDF. You must have all of these functions installed to run getxcatdocs. See: http://sourceforge.net/p/xcat/wiki/Editing_and_Downloading_xCAT_Documentation/#converting-wiki-pages-to-html-and-pdfs
-?|-h|--help
Display usage message.
-v|--verbose
Run the command in verbose mode.
-c|--continue
If a previous run of this command failed (which often happens if you lose your network connection), continue processing using files already downloaded to your markdown directory.
-d|--doc single_doc
Run this command for a single document only. If you get errors about Official-xcat-doc.png not found, either download this image directly from http://sourceforge.net/p/xcat/wiki/XCAT_Documentation/attachment/Official-xcat-doc.png or run getxcatdocs -d XCAT_Documentation first.
To download/convert the documentation and put it in ~/tmp:
getxcatdocs ~/tmp
/opt/xcat/bin/getxcatdocs
groupfiles4dsh.1¶
groupfiles4dsh - Builds a directory of files for each defined nodegroup in xCAT.
This tool will build a directory of files, one for each defined nodegroup in xCAT. The file will be named the nodegroup name and contain a list of nodes that belong to the nodegroup. The file can be used as input to the AIX dsh command. The purpose of this tool is to allow backward compatiblity with scripts that were created using the AIX or CSM dsh command
Reference: man dsh.
-h Display usage message.
-v Command Version.
-p Path to the directory to create the nodegroup files (must exist).
- To create the nodegroup files in directory /tmp/nodegroupfiles, enter:
groupfiles4dsh -p /tmp/nodegroupfiles
To use with dsh:
export DSH_CONTEXT=DSH ( default unless CSM is installed)
export DSH_NODE_RSH=/bin/ssh (default is rsh)
export DSH_NODEGROUP_PATH= /tmp/nodegroupfiles
dsh -N all date (where all is a group defined in xCAT)
dsh -a date (will look in all nodegroupfiles and build a list of all nodes)
/opt/xcat/share/xcat/tools/groupfiles4dsh
xdsh(1)|xdsh.1
imgcapture.1¶
imgcapture - Captures an image from a Linux diskful node and create a diskless or diskful image on the management node.
imgcapture node -t | --type diskless [-p | --profile profile] [-o | --osimage osimage] [-i nodebootif] [-n nodenetdrivers] [-V | --verbose]
imgcapture node -t | --type sysclone -o | --osimage osimage [-V | --verbose]
imgcapture [-h | --help] | [-v | --version]
The imgcapture command will capture an image from one running diskful Linux node and create a diskless or diskful image for later use.
The node should be one diskful Linux node, managed by the xCAT MN, and the remote shell between MN and the node should have been configured. AIX is not supported.
The imgcapture command supports two image types: diskless and sysclone. For the diskless type, it will capture an image from one running diskful Linux node, prepares the rootimg directory, kernel and initial rmadisks for the liteimg/packimage command to generate the statelite/stateless rootimg. For the sysclone type, it will capture an image from one running diskful Linux node, create an osimage which can be used to clone other diskful Linux nodes.
The diskless type:
The osimage is pre-defined, the attributes of osimage will be used to capture and prepare the root image. The osver, arch and profile attributes for the stateless/statelite image to be created are duplicated from the node‘s attribute. If the -p|--profile profile option is specified, the image will be created under “/<installroot>/netboot/<osver>/<arch>/<profile>/rootimg”.
The default files/directories excluded in the image are specified by /opt/xcat/share/xcat/netboot/<os>/<profile>.<osver>.<arch>.imgcapture.exlist; also, you can put your customized file (<profile>.<osver>.<arch>.imgcapture.exlist) to /install/custom/netboot/<osplatform>. The directories in the default .imgcapture.exlist file are necessary to capture image from the diskful Linux node managed by xCAT, please don’t remove it.
The image captured will be extracted into the /<installroot>/netboot/<osver>/<arch>/<profile>/rootimg directory.
After the imgcapture command returns without any errors, you can customize the rootimg and run the liteimg/packimage command with the options you want.
The sysclone type:
xCAT leverages the Open Source Tool - Systemimager to capture the osimage from the node, and put it into /<installroot>/sysclone/images directory.
The imgcapture command will create the osimage definition after the image is captured successfully, you can use this osimage and nodeset command to clone diskful nodes.
-t | --type
Specify the osimage type you want to capture, two types are supported: diskless and sysclone.
-p|--profile profile
Assign profile as the profile of the image to be created.
-o|--osimage osimage
The osimage name.
-i nodebootif
The network interface the diskless node will boot over (e.g. eth0), which is used by the genimage command to generate initial ramdisks.
This is optional.
-n nodenetdrivers
The driver modules needed for the network interface, which is used by the genimage command to generate initial ramdisks.
This is optional. By default, the genimage command can provide drivers for the following network interfaces:
For x86 or x86_64 platform:
tg3 bnx2 bnx2x e1000 e1000e igb m1x_enFor ppc64 platform:
e1000 e1000e igb ibmveth eheaFor S390x:
qdio ccwgroupIf the network interface is not in the above list, you’d better specify the driver modules with this option.
-h|--help
Display the usage message.
-v|--version
Display the version.
-V|--verbose
Verbose output.
node1 is one diskful Linux node, which is managed by xCAT.
- In order to capture and prepare the diskless root image, run the following command:
imgcapture node1 -t diskless
- In order to capture and prepare the diskless root image with hpc as profile, run the command:
imgcapture node1 -t diskless -p hpc
- In order to capture and prepare the diskless root image: its profile is hpc, and the network interface the diskless node will boot over is eth0, the driver modules for this network interface is e1000e.
imgcapture node1 -t diskless -p hpc -i eth0 -n e1000e
- There’s one pre-defined osimage. In order to capture and prepare the diskless root image for osimage, run the command:
imgcapture node1 -t diskless -o osimage
- In order to capture the diskful image from node1 and create the osimage img1, run the command:
imgcapture node1 -t sysclone -o img1
/opt/xcat/bin/imgcapture
genimage(1)|genimage.1, imgimport(1)|imgimport.1, imgexport(1)|imgexport.1, packimage(1)|packimage.1, liteimg(1)|liteimg.1, nodeset(8)|nodeset.8
imgexport.1¶
imgexport - Exports an xCAT image.
imgexport [-h| --help]
imgexport image_name [destination] [[-e|--extra file:dir] ... ] [-p|--postscripts node_name] [-v|--verbose]
The imgexport command will export an image that is being used by xCAT. To export images, you must have the images defined in the osimage table. All the columns in the osimage and linuximage tables will be exported. If kits are used in statefull or stateless images, kit, kitcomponent and kitrepo tables will be exported. In addition, the following files will also be exported.
- For statefull:
- x.pkglist x.otherpkgs.pkglist x.tmpl x.synclist kits related files
- For stateless:
- kernel initrd.gz rootimg.gz x.pkglist x.otherpkgs.pkglist x.synclist x.postinstall x.exlist kits related files
- For statelite:
- kernel initrd.gz root image tree x.pkglist x.synclist x.otherpkgs.pkglist x.postinstall x.exlist
where x is the name of the profile.
Any files specified by the -e flag will also be exported. If -p flag is specified, the names of the postscripts and the postbootscripts for the given node will be exported. The postscripts themsleves need to be manualy exported using -e flag.
For statelite, the litefile table settings for the image will also be exported. The litetree and statelite tables are not exported.
-e|--extra srcfile:destdir Pack up extra files. If destdir is omitted, the destination directory will be the same as the source directory.
-h|--help Display usage message.
-p|--postscripts node_name Get the names of the postscripts and postbootscripts for the given node and pack them into the image.
-v|--verbose Verbose output.
image_name The name of the image. Use lsdef -t osimage to find out all the image names.
destination The output bundle file name.
- Simplest way to export an image. If there is an image in the osimage table named ‘foo’, then run:
imgexport foo
foo.tgz will be built in the current working directory. Make sure that you have enough space in the directory that you are in to run imgexport if you have a big image to tar up.
- To include extra files with your image:
imgexport Default_Stateless_1265981465 foo.tgz -e /install/postscripts/myscript1 -e /tmp/mydir:/usr/mydir
In addition to all the default files, this will export /install/postscripts/myscript1 and the whole directory /tmp/dir into the file called foo.tgz. And when imgimport is called /install/postscripts/myscript1 will be copied into the same directory and /tmp/mydir will be copied to /usr/mydir.
- To include postscript with your image:
imgexport Default_Stateless_1265981465 foo.tgz -p node1 -e /install/postscripts/myscript1
The postscripts and the postbootscripts names specified in the postscripts table for node1 will be exported into the image. The postscript myscript1 will also be exported.
/opt/xcat/bin/imgexport
imgimport(1)|imgimport.1
imgimport.1¶
imgimport - Imports an xCAT image or configuration file into the xCAT tables so that you can immediately begin deploying with it.
imgimport [-h|--help]
imgimport bundle_file_name [-p|-**-postscripts nodelist] [-f|--profile new_profile] [-v|-**-verbose]>
The imgimport command will import an image that has been exported by imgexport from xCAT. This is the easiest way to transfer/backup/, change or share images created by xCAT whether they be stateless or stateful. The bundle file will be unpacked in the current working directory. The xCAT configuration such as osimage and linuximage tables will then be updated.
- For statefull, the following files will be copied to the appropriate directories.
- x.pkglist x.otherpkgs.pkglist x.tmpl x.synclist kits related files
- For stateless, the following files will be copied to the appropriate directories.
- kernel initrd.gz rootimg.gz x.pkglist x.otherpkgs.pkglist x.synclist x.postinstall x.exlist kits related files
- For statelite, the following files will be copied to the appropriate directories.
- kernel initrd.gz root image tree x.pkglist x.synclist x.otherpkgs.pkglist x.postinstall x.exlist
where x is the profile name.
Any extra files, included by –extra flag in the imgexport command, will also be copied to the appropriate directories.
For statelite, the litefile table will be updated for the image. The litetree and statelite tables are not imported.
If -p flag is specified, the postscripts table will be updated with the postscripts and the postbootscripts names from the image for the nodes given by this flag.
If -f flag is not specified, all the files will be copied to the same directories as the source. If it is specified, the old profile name x will be changed to the new and the files will be copied to the appropriate directores for the new profiles. For example, /opt/xcat/share/xcat/netboot/sles/x.pkglist will be copied to /install/custom/netboot/sles/compute_new.pkglist and /install/netboot/sles11/ppc64/x/kernel will be copied to /install/netboot/sles11/ppc64/compute_new/kernel. This flag is commonly used when you want to copy the image on the same xCAT mn so you can make modification on the new one.
After this command, you can run the nodeset command and then start deploying the nodes. You can also choose to modify the files and run the following commands before the node depolyment.
- For statefull:
- nodeset
- For stateless:
- genimage packimage nodeset
- For statelite
- genimage liteimg nodeset
-f|--profile new_prof Import the image with a new profile name.
-h|--help Display usage message.
-p|--postscripts nodelist Import the postscripts. The postscripts contained in the image will be set in the postscripts table for nodelist.
-v|--verbose Verbose output.
- Simplest way to import an image. If there is a bundle file named ‘foo.gz’, then run:
imgimport foo.gz
- Import the image with postscript names.
imgimport foo.gz -p node1,node2
The postscripts table will be updated with the name of the postscripts and the postbootscripts for node1 and node2.
- Import the image with a new profile name
imgimport foo.gz -f compute_test
/opt/xcat/bin/imgimport
imgexport(1)|imgexport.1
liteimg.1¶
liteimg - Modify statelite image by creating a series of links.
liteimg [-h| –help]
liteimg [-v| –version]
liteimg [-o OS] [ -p profile] [-a architecture] [-t rootfstype]
liteimg imagename
This command modifies the statelite image by creating a series of links. It creates 2 levels of indirection so that files can be modified while in their image state as well as during runtime. For example, a file like <$imgroot>/etc/ntp.conf will have the following operations done to it:
* mkdir -p $imgroot/.default/etc*
* mkdir -p $imgroot/.statelite/tmpfs/etc*
* mv $imgroot/etc/ntp.conf $imgroot/.default/etc*
* cd $imgroot/.statelite/tmpfs/etc*
* ln -sf ../../../.default/etc/ntp.conf .*
* cd $imgroot/etc*
* ln -sf ../.statelite/tmpfs/etc/ntp.conf .*
When finished, the original file will reside in $imgroot/.default/etc/ntp.conf. $imgroot/etc/ntp.conf will link to $imgroot/.statelite/tmpfs/etc/ntp.conf which will in turn link to $imgroot/.default/etc/ntp.conf
Note: If you make any changes to your litefile table after running liteimg then you will need to rerun liteimg again.
imagename specifies the name of a os image definition to be used. The specification for the image is storted in the osimage table and linuximage table.
-h Display usage message.
-v Command Version.
-o Operating system (fedora8, rhel5, sles10,etc)
-p Profile (compute,service)
-a Architecture (ppc64,x86_64,etc)
-t The type of rootfs for the image (nfs, ramdisk). It is optional, nfs is used to be the rootfs type when -t is not specified.
- To lite a fedora8 image for a compute node architecture x86_64 enter:
liteimg -o fedora8 -p compute -a x86_64
/opt/xcat/bin/
This command is part of the xCAT software product.
genimage(1)|genimage.1
lsdef.1¶
lsdef - Use this command to list xCAT data object definitions.
lsdef [-h | --help] [-t object-types] [-i attr-list]
lsdef [-V | --verbose] [-l | --long] [-s | --short] [-a | --all] [-S] [-t object-types] [-o object-names] [-z | --stanza] [-i attr-list] [-c | --compress] [--osimage] [--nics] [[-w attr==val] [-w attr=~val] ...] [noderange]
This command is used to display xCAT object definitions which are stored in the xCAT database.
-a|--all
Display all definitions. For performance consideration, the auditlog and eventlog objects will not be listed. To list auditlog or eventlog objects, use lsdef -t auditlog or lsdef -t eventlog instead.
-c|--compress
Display information in compressed mode, each output line has format “<object name>: <data>”. The output can be passed to command xcoll or xdshbak for formatted output. The -c flag must be used with -i flag.
-h|--help
Display usage message.
-i attr-list
Comma separated list of attribute names to display.
-l|--long
List the complete object definition.
-s|--short
Only list the object names.
-S
List all the hidden nodes (FSP/BPA nodes) with other ones.
noderange
A set of comma delimited node names and/or group names. See the “noderange” man page for details on supported formats.
-o object-names
A set of comma delimited object names.
--osimage
Show all the osimage information for the node.
--nics
Show the nics configuration information for the node.
-t object-types
A set of comma delimited object types. Use the help option to get a list of valid objects.
-V|--verbose
Verbose mode.
-w attr==val -w attr=~val ...
Use one or multiple -w flags to specify the selection string that can be used to select objects. The operators ==, !=, =~ and !~ are available. Use the help option to get a list of valid attributes for each object type.
- Operator descriptions:
- == Select nodes where the attribute value is exactly this value. != Select nodes where the attribute value is not this specific value. =~ Select nodes where the attribute value matches this regular expression. !~ Select nodes where the attribute value does not match this regular expression.
Note: if the “val” fields includes spaces or any other characters that will be parsed by shell, the “attr<operator>val” needs to be quoted. If the operator is ”!~”, the “attr<operator>val” needs to be quoted using single quote.
-z|--stanza
Display output in stanza format. See the xcatstanzafile man page for details on using xCAT stanza files.
To display a description of all the valid attributes that could be used when defining an xCAT node.
lsdef -t node -h
To get a list of all the objects that have been defined.
lsdef OR lsdef -a
To get all the attributes of the node1
lsdef node1 OR lsdef -t node node1 OR lsdef -t node -o node1
To get the object name of node1 instead of all the attributes
lsdef -s node1
To get a list of all the network definitions.
lsdef -t network
To get a complete listing of all network definitions.
lsdef -l -t network
To list the whole xCAT database and write it to a stanza file. (backup database)
lsdef -a -l -z > mydbstanzafile
To list the MAC and install adapter name for each node.
lsdef -t node -i mac,installnic
To list an osimage definition named “aix53J”.
lsdef -t osimage -l -o aix53J
To list all node definitions that have a status value of “booting”.
lsdef -t node -w status==booting
To list all the attributes of the group “service”.
lsdef -l -t group -o service
To list all the attributes of the nodes that are members of the group “service”.
lsdef -t node -l service
To get a listing of object definitions that includes information about what xCAT database tables are used to store the data.
lsdef -V -l -t node -o node01
To list the hidden nodes that can’t be seen with other flags. The hidden nodes are FSP/BPAs.
lsdef -S
To list the nodes status and use xcoll to format the output.
lsdef -t node -i status -c | xcoll
To display the description for some specific attributes that could be used when defining an xCAT node.
lsdef -t node -h -i profile,pprofile
To display the nics configuration information for node cn1.
lsdef cn1 --nics
/opt/xcat/bin/lsdef
This command is part of the xCAT software product.
mkdef(1)|mkdef.1, chdef(1)|chdef.1, rmdef(1)|rmdef.1, xcatstanzafile(5)|xcatstanzafile.5
lsflexnode.1¶
lsflexnode - Display the information of flexible node
IBM BladeCenter HX5 offers flexibility ideal that the blades can be combined together for scalability.
There are several concepts to support the HX5 multiple blades combination:
Complex: Multiple blades which combined by a scalability card is a complex.
Parition: A logic concept which containing part of the Blade slot node in a complex. Each partition can map to a system to install Operating System. Each partition could have 1HX5, 1HX5+1MD or 2HX5+2MD. (MD is the Memory Drawer)
Blade slot node: The physical blade which installed in the slot of a chassis. It can be a HX5 or MD.
A Complex will be created automatically when a multiple blades combination is installed. In this Complex, every blade belongs to it is working as a Blade slot node.
A Partition can be created base on the Complex, each Partition can have one or multiple Blade slot node.
The noderange in the SYNOPSIS can be a AMM node or a blade node.
The meaning of attributes which displayed by the lsflexnode. The word ‘node’ in this section means Blade slot node.
Complex
The unique numeric identifier for a complex installed in the chassis.
Partition number
The number of partitions currently defined for this complex.
Complex node number
The number of nodes existing in this complex, regardless of their assignment to any given partition.
Partition
The unique numeric identifier for a partition defined within a complex installed in the chassis.
Partition Mode
The currently configured mode of this partition. It can be ‘partition’ or ‘standalone’.
Partition node number
The number of nodes currently defined for this partition.
Partition status
The current power status of this partition when the partition has a valid partition configuration. It can be ‘poweredoff’, ‘poweredon’, ‘resetting’ or ‘invalid’.
Node
The unique numeric identifier for this node, unique within the partition. If this node does not belong to a partition, the slot number will be displayed.
Node state
The physical power state of this node. It can be ‘poweredoff’, ‘poweredon’ or ‘resetting’.
Node slot
The base slot number where the node exists in the chassis.
Node resource
A string providing a summary overview of the resources provided by this node. It includes the CPU number, CPU frequency and Memory size.
Node type
The general categorization of the node. It can be ‘processor’, ‘memory’ or ‘io’.
Node role
Indicates if the node is assigned to a partition, and if so, provides an indication of whether the node is the primary node of the partition or not.
Flexnode state
The state of a flexible node. It is the state of the partition which this node belongs to. If this node does NOT belong to a partition, the value should be ‘invalid’.
It can be ‘poweredoff’, ‘poweredon’, ‘resetting’ or ‘invalid’.
Complex id
The identifier of the complex this node belongs to.
Partition id
The identifier of the partition this node belongs to.
1
Display all the Complex, Partition and Blade slot node which managed by a AMM.
lsflexnode amm1The output:
amm1: Complex - 24068 amm1: ..Partition number - 1 amm1: ..Complex node number - 2 amm1: ..Partition = 1 amm1: ....Partition Mode - partition amm1: ....Partition node number - 1 amm1: ....Partition status - poweredoff amm1: ....Node - 0 (logic id) amm1: ......Node state - poweredoff amm1: ......Node slot - 14 amm1: ......Node type - processor amm1: ......Node resource - 2 (1866 MHz) / 8 (2 GB) amm1: ......Node role - secondary amm1: ..Partition = unassigned amm1: ....Node - 13 (logic id) amm1: ......Node state - poweredoff amm1: ......Node slot - 13 amm1: ......Node type - processor amm1: ......Node resource - 2 (1866 MHz) / 8 (2 GB) amm1: ......Node role - unassigned
2
Display a flexible node.
lsflexnode blade1The output:
blade1: Flexnode state - poweredoff blade1: Complex id - 24068 blade1: Partition id - 1 blade1: Slot14: Node state - poweredoff blade1: Slot14: Node slot - 14 blade1: Slot14: Node type - processor blade1: Slot14: Node resource - 2 (1866 MHz) / 8 (2 GB) blade1: Slot14: Node role - secondary
/opt/xcat/bin/lsflexnode
mkflexnode(1)|mkflexnode.1, rmflexnode(1)|rmflexnode.1
lshwconn.1¶
lshwconn - Use this command to display the connection status for CEC and Frame nodes.
lshwconn [-h| --help]
lshwconn [-v| --version]
lshwconn [-V| --verbose] noderange
lshwconn noderange -T tooltype
This command is used to display the connection status for CEC and Frame node.
-h|--help
Display usage message.
-V|--verbose
Verbose output.
-T
The tooltype is used to communicate to the CEC/Frame. The value could be lpar or fnm. The tooltype value lpar is for xCAT and fnm is for CNM.
To display connection status for all CEC nodes in node group CEC:
lshwconn cec
Output is similar to:
cec1: ipaddr=192.168.200.245,alt_ipaddr=unavailable,state=Connected cec2: Connection not found
To display connection status for Frame node frame1:
lshwconn frame1
Output is similar to:
frame1: side=a,ipaddr=192.168.200.247,alt_ipaddr=unavailable,state=Connected frame1: side=b,ipaddr=192.168.200.248,alt_ipaddr=unavailable,state=Connected
To display connection status for all CEC nodes in node group CEC to hardware server, and using lpar tooltype:
lshwconn cec -T lpar
Output is similar to:
cec1: sp=primary,ipadd=40.3.7.1,alt_ipadd=unavailable,state=LINE UP cec2: Connection not found
$XCATROOT/bin/lshwconn
(The XCATROOT environment variable is set when xCAT is installed. The default value is “/opt/xcat”.)
This command is part of the xCAT software product.
rmhwconn(1)|rmhwconn.1, mkhwconn(1)|mkhwconn.1
lskit.1¶
lskit - Lists information for one or more Kits.
- lskit [-V | --verbose]
- [-F | --framework kitattr_names] [-x | --xml | --XML] [-K | --kitattr kitattr_names] [-R | --repoattr repoattr_names] [-C | --compattr compattr_names] [kit_names]
lskit [-? | -h | --help | -v | --version]
lskit [-F | --framework kit_path_name]
The lskit command is used to list information for one or more kits. A kit is a special kind of package that is used to install a software product on one or more nodes in an xCAT cluster.
Note: The xCAT support for Kits is only available for Linux operating systems.
The lskit command outputs the following info for each kit: the kit’s basic info, the kit’s repositories, and the kit’s components. The command outputs the info in two formats: human-readable format (default), and XML format. Use the -x option to view the info in XML format.
Input to the command can specify any number or combination of the input options.
-F|--framework kit_path_name
Use this option to display the framework values of the specified Kit tarfile. This information is retreived directly from the tarfile and can be done before the Kit has been defined in the xCAT database. This option cannot be combined with other options.
-K|--kitattr kitattr_names
Where kitattr_names is a comma-delimited list of kit attribute names. The names correspond to attribute names in the kit table. The lskit command will only display the specified kit attributes.
-R|--repoattr repoattr_names
Where repoattr_names is a comma-delimited list of kit repository attribute names. The names correspond to attribute names in the kitrepo table. The lskit command will only display the specified kit repository attributes.
-C|--compattr compattr_names
where compattr_names is a comma-delimited list of kit component attribute names. The names correspond to attribute names in the kitcomponent table. The lskit command will only display the specified kit component attributes.
kit_names
is a comma-delimited list of kit names. The lskit command will only display the kits matching these names.
-x|--xml|--XML
Need XCATXMLTRACE=1 env when using -x|–xml|–XML, for example: XCATXMLTRACE=1 lskit -x testkit-1.0.0 Return the output with XML tags. The data is returned as:
- <data>
- <kitinfo>
- ...
</kitinfo>
</data> ... <data>
- <kitinfo>
- ...
</kitinfo>
</data>
- Each <kitinfo> tag contains info for one kit. The info inside <kitinfo> is structured as follows:
- The <kit> sub-tag contains the kit’s basic info. The <kitrepo> sub-tags store info about the kit’s repositories. The <kitcomponent> sub-tags store info about the kit’s components.
- The data inside <kitinfo> is returned as:
- <kitinfo>
- <kit>
- ...
</kit>
<kitrepo> ... </kitrepo> ... <kitcomponent> ... </kitcomponent> ... </kitinfo>
-V|--verbose
Display additional progress and error messages.
-v|--version
Command Version.
-?|-h|--help
Display usage message.
To list all kits, enter:
lskitTo list the kit “kit-test1-1.0-Linux”, enter:
lskit kit-test1-1.0-Linux
To list the kit “kit-test1-1.0-Linux” for selected attributes, enter:
lskit -K basename,description -R kitreponame -C kitcompname kit-test1-1.0-Linux
To list the framework value of a Kit tarfile.
lskit -F /myhome/mykits/pperte-1.3.0.2-0-x86_64.tar.bz2 Extracting the kit.conf file from /myhome/mykits/pperte-1.3.0.2-0-x86_64.tar.bz2. Please wait. kitframework=2 compatible_kitframeworks=0,1,2
To list kit “testkit-1.0-1” with XML tags, enter:
XCATXMLTRACE=1 lskit -x testkit-1.0-1
/opt/xcat/bin/lskit
lskitcomp(1)|lskitcomp.1, lskitdeployparam(1)|lskitdeployparam.1, addkit(1)|addkit.1, rmkit(1)|rmkit.1, addkitcomp(1)|addkitcomp.1, rmkitcomp(1)|rmkitcomp.1
lskitcomp.1¶
lskitcomp - Used to list information for one or more kit components.
- lskitcomp [-V | --verbose]
- [-x | --xml | --XML] [-C | --compattr compattr_names] [-O | --osdistro os_distro] [-S | --serverrole server_role] [kitcomp_names]
lskitcomp [-? | -h | --help | -v | --version]
The lskitcomp command is used to list information for one or more kit components. A kit is made up of one or more kit components. Each kit component is a meta package used to install a software product component on one or more nodes in an xCAT cluster.
The lskitcomp command outputs the kit component info in two formats: human-readable format (default), and XML format. Use the -x option to view the info in XML format.
Input to the command can specify any number or combination of the input options.
Note: The xCAT support for Kits is only available for Linux operating systems.
-C|--compattr compattr_names
where compattr_names is a comma-delimited list of kit component attribute names. The names correspond to attribute names in the kitcomponent table. The lskitcomp command will only display the specified kit component attributes.
-O|--osdistro os_distro
where os_distro is the name of an osdistro in osdistro table. The lskitcomp command will only display the kit components matching the specified osdistro.
-S|--serverrole server_role
where server_role is the name of a server role. The typical server roles are: mgtnode, servicenode, computenode, loginnode, storagennode. The lskitcomp command will only display the kit components matching the specified server role.
kitcomp_names
is a comma-delimited list of kit component names. The lskitcomp command will only display the kit components matching the specified names.
-x|--xml|--XML
Need XCATXMLTRACE=1 env when using -x|–xml|–XML. Return the output with XML tags. The data is returned as:
- <data>
- <kitinfo>
- ...
</kitinfo>
</data> ... <data>
- <kitinfo>
- ...
</kitinfo>
</data>
- Each <kitinfo> tag contains info for a group of kit compoonents belonging to the same kit. The info inside <kitinfo> is structured as follows:
- The <kit> sub-tag contains the kit’s name. The <kitcomponent> sub-tags store info about the kit’s components.
- The data inside <kitinfo> is returned as:
- <kitinfo>
- <kit>
- ...
</kit>
<kitcomponent> ... </kitcomponent> ... </kitinfo>
-V|--verbose
Display additional progress and error messages.
-v|--version
Command Version.
-?|-h|--help
Display usage message.
To list all kit components, enter:
lskitcompTo list the kit component “comp-server-1.0-1-rhels-6-x86_64”, enter:
lskitcomp comp-server-1.0-1-rhels-6-x86_64
To list the kit component “comp-server-1.0-1-rhels-6-x86_64” for selected kit component attributes, enter:
lskitcomp -C kitcompname,desc comp-server-1.0-1-rhels-6-x86_64
To list kit components compatible with “rhels-6.2-x86_64” osdistro, enter:
lskitcomp -O rhels-6.2-x86_64
To list kit components compatible with “rhels-6.2-x86_64” osdistro and “computenode” server role, enter:
lskitcomp -O rhels-6.2-x86_64 -S computenode
To list the kit component “testkit-compute-1.0-1-ubuntu-14.04-ppc64el” with XML tags, enter:
XCATXMLTRACE=1 lskitcomp -x testkit-compute-1.0-1-ubuntu-14.04-ppc64el
/opt/xcat/bin/lskitcomp
lskit(1)|lskit.1, lskitdeployparam(1)|lskitdeployparam.1, addkit(1)|addkit.1, rmkit(1)|rmkit.1, addkitcomp(1)|addkitcomp.1, rmkitcomp(1)|rmkitcomp.1
lskitdeployparam.1¶
lskitdeployparam - Lists the deployment parameters for one or more Kits or Kit components
- lskitdeployparam [-V | --verbose]
- [-x | --xml | --XML] [-k | --kitname kit_names] [-c | --compname comp_names]
lskitdeployparam [-? | -h | --help | -v | --version]
The lskitdeployparam command is used to list the kit deployment parameters for one or more kits, or one or more kit components. Kit deployment parameters are used to customize the installation or upgrade of kit components.
The lskitdeployparam command outputs the kit component information in two formats: human-readable format (default), and XML format. Use the -x option to view the information in XML format.
Input to the command can specify any combination of the input options.
Note: The xCAT support for Kits is only available for Linux operating systems.
-k|--kitname kit_names
Where kit_names is a comma-delimited list of kit names. The lskitdeployparam command will only display the deployment parameters for the kits with the matching names.
-c|--compname comp_names
Where comp_names is a comma-delimited list of kit component names. The lskitdeployparam command will only display the deployment parameters for the kit components with the matching names.
-x|--xml|--XML
- Return the output with XML tags. The data is returned as:
- <data>
- <kitdeployparam>
- <name>KIT_KIT1_PARAM1</name> <value>value11</value>
</kitdeployparam>
</data> <data>
- <kitdeployparam>
- <name>KIT_KIT1_PARAM2</name> <value>value12</value>
</kitdeployparam>
</data> ...
-V|--verbose
Display additional progress and error messages.
-v|--version
Command Version.
-?|-h|--help
Display usage message.
To list kit deployment parameters for kit “kit-test1-1.0-Linux”, enter:
lskitdeployparam -k kit-test1-1.0-Linux
To list kit deployment parameters for kit component “comp-server-1.0-1-rhels-6-x86_64”, enter:
lskitdeployparam -c comp-server-1.0-1-rhels-6-x86_64
/opt/xcat/bin/lskitdeployparam
lskit(1)|lskit.1, lskitcomp(1)|lskitcomp.1, addkit(1)|addkit.1, rmkit(1)|rmkit.1, addkitcomp(1)|addkitcomp.1, rmkitcomp(1)|rmkitcomp.1
lskmodules.1¶
lskmodules - list kernel driver modules in rpms or driver disk image files
- lskmodules [-V | --verbose]
- [-i | --osimage osimage_names] [-c | --kitcomponent kitcomp_names] [-o | --osdistro osdistro_names] [-u | --osdistropudate osdistroupdate_names] [-x | --xml | --XML]
lskmodules [-? | -h | --help | -v | --version]
The lskmodules command finds the kernel driver module files (*.ko) in the specified input locations, runs the modinfo command against each file, and returns the driver name and description. If -x is specified, the output is returned with XML tags.
Input to the command can specify any number or combination of the input options.
-i|--osimage osimage_names
where osimage_names is a comma-delimited list of xCAT database osimage object names. For each osimage_name, lskmodules will use the entries in osimage.driverupdatesrc for the rpms and driver disk image files to search.
-c|--kitcomponent kitcomponent_names
where kitcomponent_names is a comma-delimited list of xCAT database kitcomponent object names. For each kitcomponent_name, lskmodules will use the entries in kitcomponent.driverpacks for the rpm list and the repodir of the kitcomponent.kitreponame for the location of the rpm files to search.
-o|--osdistro osdistro_names
where osdistro_names is a comma-delimited list of xCAT database osdistro object names. For each osdistro_name, lskmodules will search each <osdistro.dirpaths>/Packages/kernel-<kernelversion>.rpm file.
-u|--osdistroupdate osdistroupdate_names
where osdistroupdate_names is a comma-delimited list of xCAT database osdistroupdate table entries. For each osdistroupdate_name, lskmodules will search the <osdistroupdate.dirpath>/kernel-<kernelversion>.rpm file.
-x|--xml|--XML
- Return the output with XML tags. The data is returned as:
- <module>
- <name> xxx.ko </name> <description> this is module xxx </description>
</module>
This option is intended for use by other programs. The XML will not be displayed. To view the returned XML, set the XCATSHOWXML=yes environment variable before running this command.
-V|--verbose
Display additional progress and error messages.
-v|--version
Command Version.
-?|-h|--help
Display usage message.
To list the kernel modules included in the driverpacks shipped with kitcomponent kit1_comp1-x86_64, enter:
lskmodules -c kit1_comp1-x86_64
lslite.1¶
lslite - Display a summary of the statelite information.
The lslite command displays a summary of the statelite information that has been defined for a noderange or an image.
-h|--help
Display usage message.
-V|--verbose
Verbose mode.
-i imagename
The name of an existing xCAT osimage definition.
noderange
A set of comma delimited node names and/or group names. See the “noderange” man page for details on additional supported formats.
To list the statelite information for an xCAT node named “node01”.
lslite node01
Output is similar to:
>>>Node: node01
Osimage: 61img
- Persistent directory (statelite table):
xcatmn1:/statelite
- Litefiles (litefile table):
tmpfs,rw /etc/adjtime tmpfs,rw /etc/lvm/.cache tmpfs,rw /etc/mtab ........
- Litetree path (litetree table):
1,MN:/etc 2,server1:/etc
To list the statelite information for an xCAT osimage named “osimage01”.
lslite -i osimage01
Output is similar to:
tmpfs,rw /etc/adjtime tmpfs,rw /etc/lvm/.cache tmpfs,rw /etc/mtab ........
/opt/xcat/bin/lslite
noderange(3)|noderange.3, tabdump(8)|tabdump.8
lsslp.1¶
lsslp - Discovers selected networked services information within the same subnet.
lsslp [-h| –help]
lsslp [-v| –version]
lsslp [noderange] [-V] [-i ip[,ip..]][-w][-r|-x|-z][-n][-s CEC|FRAME|MM|IVM|RSA|HMC|CMM|IMM2|FSP][-t tries][-I][-C counts][-T timeout][–vpdtable]
The lsslp command discovers selected service types using the -s flag. All service types are returned if the -s flag is not specified. If a specific IP address is not specified using the -i flag, the request is sent out all available network adapters. The optional -r, -x, -z and –vpdtable flags format the output. If you can’t receive all the hardware, please use -T to increase the waiting time.
NOTE: SLP broadcast requests will propagate only within the subnet of the network adapter broadcast IPs specified by the -i flag.
- noderange The nodes which the user want to discover.
- If the user specify the noderange, lsslp will just return the nodes in the node range. Which means it will help to add the new nodes to the xCAT database without modifying the existed definitions. But the nodes’ name specified in noderange should be defined in database in advance. The specified nodes’ type can be frame/cec/hmc/fsp/bpa. If the it is frame or cec, lsslp will list the bpa or fsp nodes within the nodes(bap for frame, fsp for cec). Please do not use noderange with the flag -s.
-i IP(s) the command will send out (defaults to all available adapters).
-h Display usage message.
-n Only display and write the newly discovered hardwares.
- -u Do unicast to a specified IP range. Must be used with -s and -****-range.
- The -u flag is not supported on AIX.
- --range Specify one or more IP ranges. Must be use in unicast mode.
- It accepts multiple formats. For example, 192.168.1.1/24, 40-41.1-2.3-4.1-100. If the range is huge, for example, 192.168.1.1/8, lsslp may take a very long time for node scan. So the range should be exactly specified.
-r Display Raw SLP response.
- -C The number of the expected responses specified by the user.
- When using this flag, lsslp will not return until the it has found all the nodes or time out. The default max time is 3 secondes. The user can use -T flag the specify the time they want to use. A short time will limite the time costing, while a long time will help to find all the nodes.
-T The number in seconds to limite the time costing of lsslp.
-s Service type interested in discovering.
-t Number or service-request attempts.
--vpdtable Output the SLP response in vpdtable formatting. Easy for writting data to vpd table.
-v Command Version.
-V Verbose output.
-w Writes output to xCAT database.
-x XML format.
-z Stanza formated output.
- -I Give the warning message for the nodes in database which have no SLP responses.
- Please note that this flag noly can be used after the database migration finished successfully.
- To list all discovered HMC service types in tabular format, enter:
lsslp -s HMC
Output is similar to:
device type-model serial-number ip-addresses hostname
HMC 7310CR2 103F55A 1.1.1.115 hmc01
HMC 7310CR2 105369A 3.3.3.103 hmc02
HMC 7310CR3 KPHHK24 3.3.3.154 hmc03
- list all discovered FSP service types in raw response format on subnet 30.0.0.255, enter:
lsslp -i 3.0.0.255 -s CEC -r
Output is similar to:
(type=cec-service-processor),(serial-number=10A3AEB),(machinetype-model=9117-570),(fru-serial-number=YL11C5338102),(hostname=),(frame-number=0),(cage-number=0),(ip-address=3.0.0.94,1.1.1.147),(web-url=https://3.0.0.94:473 ), (slot=1),(bpc-machinetype-model=0),(bpc-serial-number=0),(Image=fips240/b0630a_0623.240)
(type=cec-service-processor),(serial-number=10A3E2B),(machinetype-model=9117-570),(fru-serial- number=YL11C5338250),(hostname=),(frame-number=0),(cage-number=0),(ip-address=3.0.0.95,1.1.1.147), (web-url=https://3.0.0.95:473 ),(slot=1),(bpc-machinetype-model=0),(bpc-serial-number=0),(Image=fips240/b0630a_0623.240)
- To list all discovered MM service types in XML format and write the output to the xCAT database, enter:
lsslp -s MM -x -w
Output is similar to:
<Node>
<groups>mm,all</groups>
<id>00:14:5E:E0:CB:1E</id>
<mgt>blade</mgt>
<mtm>029310C</mtm>
<node>Server-029310C-SN100485A-A</node>
<nodetype>mm</nodetype>
<otherinterfaces>9.114.47.229</otherinterfaces>
<serial>100485A</serial>
</Node>
- To list all discovered service types in stanza format and write the output to the xCAT database, enter:
lsslp -z -w
Output is similar to:
- c76v1hmc02:
- objtype=node hcp=c76v1hmc02 nodetype=hmc mtm=7315CR2 serial=10407DA ip=192.168.200.125 groups=hmc,all mgt=hmc mac=00:1a:64:fb:7d:50 hidden=0
- 192.168.200.244:
- objtype=node hcp=192.168.200.244 nodetype=fsp mtm=9125-F2A serial=0262662 side=A-0 otherinterfaces=192.168.200.244 groups=fsp,all mgt=fsp id=4 parent=Server-9125-F2A-SN0262662 mac=00:1a:64:fa:01:fe hidden=1
- Server-8205-E6B-SN1074CDP:
- objtype=node hcp=Server-8205-E6B-SN1074CDP nodetype=cec mtm=8205-E6B serial=1074CDP groups=cec,all mgt=fsp id=0 hidden=0
- 192.168.200.33:
- objtype=node hcp=192.168.200.33 nodetype=bpa mtm=9458-100 serial=99201WM side=B-0 otherinterfaces=192.168.200.33 groups=bpa,all mgt=bpa id=0 mac=00:09:6b:ad:19:90 hidden=1
- Server-9125-F2A-SN0262652:
- objtype=node hcp=Server-9125-F2A-SN0262652 nodetype=frame mtm=9125-F2A serial=0262652 groups=frame,all mgt=fsp id=5 hidden=0
- To list all discovered service types in stanza format and display the IP address, enter:
lsslp -w
Output is similar to:
mm01:
objtype=node
nodetype=fsp
mtm=8233-E8B
serial=1000ECP
side=A-0
groups=fsp,all
mgt=fsp
id=0
mac=00:14:5E:F0:5C:FD
otherinterfaces=50.0.0.5
bpa01:
objtype=node
nodetype=bpa
mtm=9A01-100
serial=0P1N746
side=A-1
groups=bpa,all
mgt=bpa
id=0
mac=00:1A:64:54:8C:A5
otherinterfaces=50.0.0.1
- To list all the CECs, enter:
lsslp -s CEC
device type-model serial-number side ip-addresses hostname FSP 9117-MMB 105EBEP A-1 20.0.0.138 20.0.0.138 FSP 9117-MMB 105EBEP B-1 20.0.0.139 20.0.0.139 CEC 9117-MMB 105EBEP Server-9117-MMB-SN105EBEP
- To list all the nodes defined in database which have no SLP response.
lsslp -I
Output is similar to:
These nodes defined in database but can’t be discovered: f17c00bpcb_b,f17c01bpcb_a,f17c01bpcb_b,f17c02bpcb_a,
device type-model serial-number side ip-addresses hostname bpa 9458-100 BPCF017 A-0 40.17.0.1 f17c00bpca_a bpa 9458-100 BPCF017 B-0 40.17.0.2 f17c00bpcb_a
- To find the nodes within the user specified. Please make sure the noderange input have been defined in xCAT database.
lsslp CEC1-CEC3
or lsslp CEC1,CEC2,CEC3
device type-model serial-number side ip-addresses hostname
FSP 9A01-100 0P1P336 A-0 192.168.200.34 192.168.200.34
FSP 9A01-100 0P1P336 B-0 192.168.200.35 192.168.200.35
FSP 9A01-100 0P1P336 A-1 50.0.0.27 50.0.0.27
FSP 9A01-100 0P1P336 B-1 50.0.0.28 50.0.0.28
CEC 9A01-100 0P1P336 CEC1
FSP 8233-E8B 1040C7P A-0 192.168.200.36 192.168.200.36
FSP 8233-E8B 1040C7P B-0 192.168.200.37 192.168.200.37
FSP 8233-E8B 1040C7P A-1 50.0.0.29 50.0.0.29
FSP 8233-E8B 1040C7P B-1 50.0.0.30 50.0.0.30
CEC 8233-E8B 1040C7P CEC2
FSP 8205-E6B 1000ECP A-0 192.168.200.38 192.168.200.38
FSP 8205-E6B 1000ECP B-0 192.168.200.39 192.168.200.39
FSP 8205-E6B 1000ECP A-1 50.0.0.31 50.0.0.27
FSP 8205-E6B 1000ECP B-1 50.0.0.32 50.0.0.28
CEC 8205-E6B 1000ECP CEC3
- To list all discovered CMM in stanza format, enter: lsslp -s CMM -m -z
- e114ngmm1:
- objtype=node mpa=e114ngmm1 nodetype=cmm mtm=98939AX serial=102537A groups=cmm,all mgt=blade hidden=0 otherinterfaces=70.0.0.30 hwtype=cmm
- To use lsslp unicast, enter: lsslp -u -s CEC –range 40-41.1-2.1-2.1-2
/opt/xcat/bin/lsslp
rscan(1)|rscan.1
lstree.1¶
lstree - Display the tree of service node hierarchy, hardware hierarchy, or VM hierarchy.
The lstree command can display the tree of service node hierarchy for the xCAT nodes which have service node defined or which are service nodes, display the tree of hardware hierarchy only for the physical objects, display the tree of VM hierarchy for the xCAT nodes which are virtual machines or which are the hosts of virtual machines. If a noderange is specified, only show the part of the hierarchy that involves those nodes. For ZVM, we only support to disply VM hierarchy. By default, lstree will show both the hardware hierarchy and the VM hierarchy for all the nodes.
-h|--help
Display usage message.
-s|-- servicenode
Show the tree of service node hierarchy.
-H|--hardwaremgmt
Show the tree of hardware hierarchy.
--v|--virtualmachine
Show the tree of VM hierarchy.
noderange
A set of comma delimited node names and/or group names. See the “noderange” man page for details on additional supported formats.
To display the tree of service node hierarchy for all the nodes.
lstree -s
Output is similar to:
To display the tree of service node hierarchy for service node “mysn01”.
lstree -s mysn01
Output is similar to:
To display the tree of hardware hierarchy for all the nodes.
lstree -H
Output is similar to:
To display the tree of hardware hierarchy for HMC “myhmc01”.
lstree -H myhmc01
Output is similar to:
To display the tree of VM hierarchy for all the nodes.
lstree -v
Output is similar to:
To display the tree of VM hierarchy for the node “x3650n01”.
lstree -v x3650n01
Output is similar to:
To display both the hardware tree and VM tree for all nodes.
lstree
Output is similar to:
- HMC: myhmc01
- Service Focal Point: myhmc02
- Management Module: mymm01
- BMC: 192.168.0.1
/opt/xcat/bin/lstree
noderange(3)|noderange.3, tabdump(8)|tabdump.8
lsve.1¶
lsve - Lists detail attributes for a virtual environment.
lsve [-t type] [-m manager] [-o object]
The lsve command can be used to list a virtual environment for ‘Data Center’, ‘Cluster’, ‘Storage Domain’, ‘Network’ and ‘Template’ objects.
The mandatory parameter -m manager is used to specify the address of the manager of virtual environment. xCAT needs it to access the RHEV manager.
The mandatory parameter -t type is used to specify the type of the target object.
Basically, lsve command supports three types of object: dc, cl, sd, nwand tpl.
The parameter -o object is used to specify which object to list. If no -o is specified, all the objects with the -t type will be displayed.
-h Display usage message.
-m Specify the manager of the virtual environment.
For RHEV, the FQDN (Fully Qualified Domain Name) of the rhev manager have to be specified.
-o The target object to display.
-t Specify the type of the target object.
- Supported types:
- dc - Data Center (For type of ‘dc’, all the elements belongs to the data
- center will be listed.)
cl - Cluster sd - Storage Domain (To get the status of Storage Doamin, show it from
data center it attached to.nw - Network tpl - Template
- To list the data center ‘Default’, enter:
lsve -t B<dc> -m <FQDN of rhev manager> -o DefaultOutput is similar to:
datacenters: [Default] description: The default Data Center state: up storageformat: v1 storagetype: nfs clusters: [Default] cpu: Intel Westmere Family description: The default server cluster memory_hugepage: true memory_overcommit: 100 storagedomains: [image] available: 55834574848 committed: 13958643712 ismaster: true status: active storage_add: <Address of storage domain> storage_format: v1 storage_path: /vfsimg storage_type: nfs type: data used: 9663676416 networks: [rhevm2] description: state: operational stp: false networks: [rhevm] description: Management Network state: operational stp: false templates: [Blank] bootorder: hd cpucore: 1 cpusocket: 1 creation_time: 2008-04-01T00:00:00.000-04:00 display: spice memory: 536870912 state: ok stateless: false type: desktop
- To list the cluster ‘Default’, enter:
lsve -t B<cl> -m <FQDN of rhev manager> -o DefaultOutput is similar to:
cpu: Intel Westmere Family description: The default server cluster memory_hugepage: true memory_overcommit: 10
- To list the Storage Domain ‘image’, enter:
lsve -t B<sd> -m <FQDN of rhev manager> -o image
- Output is similar to:
- storagedomains: [image]
- available: 55834574848 committed: 13958643712 ismaster: true status: storage_add: <Address of storage domain> storage_format: v1 storage_path: /vfsimg storage_type: nfs type: data used: 9663676416
- To list the network ‘rhevm’, enter:
lsve -t B<nw> -m <FQDN of rhev manager> -o rhevmOutput is similar to:
networks: [rhevm] description: Management Network state: operational stp: false
- To list the template ‘tpl01’, enter:
lsve -t tpl -m <FQDN of rhev manager> -o tpl01Output is similar to:
templates: [tpl01] bootorder: network cpucore: 2 cpusocket: 2 creation_time: 2012-08-22T23:52:35.953-04:00 display: vnc memory: 1999634432 state: ok stateless: false type: server
/opt/xcat/bin/lsve
cfgve(1)|cfgve.1
lsvlan.1¶
lsvlan - It lists the existing vlans for the cluster.
The lsvlan command lists all the vlans for the cluster. If vlanid is specifined it will list more details about this vlan including the nodes in the vlan.
vlanid is a unique vlan number. If it is omitted, all vlans will be listed.
To list all the vlans in the cluster
lsvlan- Output is similar to:
- vlan 3:
subnet 10.3.0.0 netmask 255.255.0.0
vlan 4: subnet 10.4.0.0 netmask 255.255.0.0
TO list the details for vlan3
lsvlan 3
- Output is similar to:
- vlan 3
subnet 10.3.0.0 netmask 255.255.0.0
hostname ip address node vm host v3n1 10.3.0.1 c68m4hsp06 v3n2 10.3.0.2 x3455n01 v3n3 10.3.0.3 x3650n01 v3n4 10.3.0.4 x3650n01kvm1 x3650n01 v3n5 10.3.0.5 x3650n01kvm2 x3650n01
/opt/xcat/bin/lsvlan
mkvlan(1)|mkvlan.1, rmvlan(1)|rmvlan.1, chvlan(1)|chvlan.1
lsvm.1¶
lsvm - Lists partition profile information for HMC-, DFM-, IVM-, KVM-, Vmware- and zVM-managed nodes. For Power 775, it lists the LPARs’ I/O slots information and CEC configuration.
lsvm [-h| --help]
lsvm [-v| --version]
lsvm [-V| --verbose] noderange
lsvm [-a| --all] noderange
lsvm noderange
The lsvm command lists all partition profiles defined for the partitions specified in noderange. If noderange is a CEC, all the partitions associated with that CEC are displayed.
For Power 775(use option –p775 to specify), lsvm lists all partition I/O slots information for the partitions specified in noderange. If noderange is a CEC, it gets the CEC’s pump mode value, octant’s memory interleaving value, the all the octants configure value, and all the I/O slots information.
For DFM-managed (short for Direct FSP Management mode) normal power machine, lsvm lists the processor, memory, physical I/O slots, hugepage and BSR info for the specified partitions or CEC.
- The pump mode value has the valid options:
- 1 - Node Pump Mode 2 - Chip Pump Mode
- The Memory Interleaving Mode has 3 valid options:
- 0 - not Applicable 1 - interleaved 2 - non-interleaved
- More information about this part, refer to the section Using the *vm commands to define partitions in xCAT DFM in the doc below.
- XCAT_Power_775_Hardware_Management
The virtual machines that defined in the hypervisor noderange will be displayed. noderange only can be hypervisor. The type of the hypervisor should be set: hypervisor.type before running the lsvm.
Note: Only the virtual machine which is in power on state can be listed by lsvm command.
Show the directory entry for a given virtual machine.
-h
Display usage message.
-v
Command Version.
-V
Verbose output.
-a
List all the profiles for one partition
--p775
Specify the operation is for Power 775 machines.
-l
Show lparnames for lpars. It shall work with option --p775.
- To list all partition profiles defined for HMC-managed partition lpar3, enter:
lsvm lpar3
Output is similar to:
lpar3: name=lpar3,lpar_name=lpar3,lpar_id=4,lpar_env=aixlinux,all_resources=0,min_mem=512, desired_mem=2048, max_mem=3072,min_num_huge_pages=0,desired_num_huge_pages=0,max_num_huge_pages=0,proc_mode=shared, min_proc_units=0.5,desired_proc_units=0.5,max_proc_units=0.5,min_procs=1,desired_procs=1,max_procs=1, sharing_mode=uncap,uncap_weight=128,shared_proc_pool_id=0,shared_proc_pool_name=DefaultPool,io_slots=none, lpar_io_pool_ids=none,max_virtual_slots=10, "virtual_serial_adapters=1/server/1/any//any/1,0/server/1/any//any/1", virtual_scsi_adapters=2/client/1/p6vios/4/1,virtual_eth_adapters=3/0/1//0/1,hca_adapters=none,boot_mode=norm,conn_monitoring=0,auto_start=0,power_ctrl_lpar_ids=none,work_group_id=none,redundant_err_path_reporting=0, bsr_arrays=0,lhea_logical_ports=none,lhea_capabilities=none,lpar_proc_compat_mode=default,electronic_err_reporting=null
2.To list all IVM-managed partitions associated with CEC cec01, enter:
lsvm cec01
g Output is similar to:
cec01: name=10-B7D1G,lpar_name=10-B7D1G,lpar_id=1,os_type=vioserver,all_resources=0,min_mem=512, desired_mem=2048,max_mem=2048,proc_mode=shared,min_proc_units=0.10,desired_proc_units=0.40, max_proc_units=4.00,min_procs=1,desired_procs=4,max_procs=4,sharing_mode=uncap,uncap_weight=128, "io_slots=21010002/none/0,21010003/none/0,21010004/none/0,21020003/none/0,21020004/none/0,21030003/none/0,21030004/none/0,21040003/none/0,21040004/none/0",lpar_io_pool_ids=none,max_virtual_slots=48, "virtual_serial_adapters=0/server/1/any//any/1,1/server/1/any//any/1,10/client/0/2/lp2/0/0,12/client/0/3/lp3/0/0,14/client/0/4/lp4/0/0","virtual_scsi_adapters=11/server/2/lp2/2/0,13/server/3/lp3/2/0,15/server/4/lp4/2/0","virtual_eth_adapters=3/0/1//1/0,4/0/2//1/0,5/0/3//1/0,6/0/4//1/0",boot_mode=norm,conn_monitoring=0,auto_start=0,power_ctrl_lpar_ids=none
name=lp2,lpar_name=lp2,lpar_id=2,os_type=aixlinux,all_resources=0,min_mem=128,desired_mem=1024,max_mem=1024,proc_mode=shared,min_proc_units=0.10,desired_proc_units=0.10,max_proc_units=4.00,min_procs=1,desired_procs=1,max_procs=4,sharing_mode=uncap,uncap_weight=128,io_slots=none,lpar_io_pool_ids=none,max_virtual_slots=6, "virtual_serial_adapters=0/server/1/any//any/1,1/server/1/any//any/1",virtual_scsi_adapters=2/client/1/10-7D1G/11/1,virtual_eth_adapters=4/0/1//0/0,boot_mode=norm,conn_monitoring=0,auto_start=0,power_ctrl_lpar_ids=none
name=lp3,lpar_name=lp3,lpar_id=3,os_type=aixlinux,all_resources=0,min_mem=128,desired_mem=128,max_mem=128,proc_mode=shared,min_proc_units=0.10,desired_proc_units=0.10,max_proc_units=4.00,min_procs=1,desired_procs=1,max_procs=4,sharing_mode=uncap,uncap_weight=128,io_slots=none,lpar_io_pool_ids=none,max_virtual_slots=6, "virtual_serial_adapters=0/server/1/any//any/1,1/server/1/any//any/1",virtual_scsi_adapters=2/client/1/10-B7D1G/13/1,virtual_eth_adapters=4/0/1//0/0,boot_mode=of,conn_monitoring=0,auto_start=1, power_ctrl_lpar_ids=none
- For Power 775, to list the I/O slot information of lpar1, enter:
lsvm lpar1 --p775
Output is similar to:
1: 514/U78A9.001.0123456-P1-C17/0x21010202/2/1
1: 513/U78A9.001.0123456-P1-C15/0x21010201/2/1
1: 512/U78A9.001.0123456-P1-C16/0x21010200/2/1
To list the lparname of lpars, enter:
lsvm lpar1 -l --p775
- Output is similar to:
- lpar1: 1: 514/U78A9.001.0123456-P1-C17/0x21010202/2/1 lpar1: 1: 513/U78A9.001.0123456-P1-C15/0x21010201/2/1 lpar1: 1: 512/U78A9.001.0123456-P1-C16/0x21010200/2/1
- For Power 775, to list the I/O slot information and octant configuration of cec1, enter:
lsvm cec1 --p775
Output is similar to:
1: 514/U78A9.001.0123456-P1-C17/0x21010202/2/1
1: 513/U78A9.001.0123456-P1-C15/0x21010201/2/1
1: 512/U78A9.001.0123456-P1-C16/0x21010200/2/1
13: 537/U78A9.001.0123456-P1-C9/0x21010219/2/13
13: 536/U78A9.001.0123456-P1-C10/0x21010218/2/13
17: 545/U78A9.001.0123456-P1-C7/0x21010221/2/17
17: 544/U78A9.001.0123456-P1-C8/0x21010220/2/17
21: 553/U78A9.001.0123456-P1-C5/0x21010229/2/21
21: 552/U78A9.001.0123456-P1-C6/0x21010228/2/21
25: 569/U78A9.001.0123456-P1-C1/0x21010239/2/25
25: 561/U78A9.001.0123456-P1-C3/0x21010231/2/25
25: 560/U78A9.001.0123456-P1-C4/0x21010230/2/25
29: 568/U78A9.001.0123456-P1-C2/0x21010238/2/29
5: 521/U78A9.001.0123456-P1-C13/0x21010209/2/5
5: 520/U78A9.001.0123456-P1-C14/0x21010208/2/5
9: 529/U78A9.001.0123456-P1-C11/0x21010211/2/9
9: 528/U78A9.001.0123456-P1-C12/0x21010210/2/9
cec1: PendingPumpMode=1,CurrentPumpMode=1,OctantCount=8:
OctantID=0,PendingOctCfg=5,CurrentOctCfg=1,PendingMemoryInterleaveMode=2,CurrentMemoryInterleaveMode=2;
OctantID=1,PendingOctCfg=1,CurrentOctCfg=1,PendingMemoryInterleaveMode=2,CurrentMemoryInterleaveMode=2;
OctantID=2,PendingOctCfg=1,CurrentOctCfg=1,PendingMemoryInterleaveMode=2,CurrentMemoryInterleaveMode=2;
OctantID=3,PendingOctCfg=1,CurrentOctCfg=1,PendingMemoryInterleaveMode=2,CurrentMemoryInterleaveMode=2;
OctantID=4,PendingOctCfg=1,CurrentOctCfg=1,PendingMemoryInterleaveMode=2,CurrentMemoryInterleaveMode=2;
OctantID=5,PendingOctCfg=1,CurrentOctCfg=1,PendingMemoryInterleaveMode=2,CurrentMemoryInterleaveMode=2;
OctantID=6,PendingOctCfg=1,CurrentOctCfg=1,PendingMemoryInterleaveMode=2,CurrentMemoryInterleaveMode=2;
OctantID=7,PendingOctCfg=1,CurrentOctCfg=1,PendingMemoryInterleaveMode=2,CurrentMemoryInterleaveMode=2;
To list the lparname of lpars, enter:
lsvm cec1 -l --p775
Output is similar to:
lpar1: 1: 514/U78A9.001.0123456-P1-C17/0x21010202/2/1: 32: 0/3/3
lpar1: 1: 513/U78A9.001.0123456-P1-C15/0x21010201/2/1: 32: 0/3/3
lpar1: 1: 512/U78A9.001.0123456-P1-C16/0x21010200/2/1: 32: 0/3/3
lpar13: 13: 537/U78A9.001.0123456-P1-C9/0x21010219/2/13: 32: 0/3/3
lpar13: 13: 536/U78A9.001.0123456-P1-C10/0x21010218/2/13: 32: 0/3/3
lpar17: 17: 545/U78A9.001.0123456-P1-C7/0x21010221/2/17: 32: 0/0/0
lpar17: 17: 544/U78A9.001.0123456-P1-C8/0x21010220/2/17: 32: 0/0/0
lpar21: 21: 553/U78A9.001.0123456-P1-C5/0x21010229/2/21: 32: 0/0/0
lpar21: 21: 552/U78A9.001.0123456-P1-C6/0x21010228/2/21: 32: 0/0/0
lpar24: 25: 569/U78A9.001.0123456-P1-C1/0x21010239/2/25: 32: 0/0/0
lpar25: 25: 561/U78A9.001.0123456-P1-C3/0x21010231/2/25: 32: 0/0/0
lpar25: 25: 560/U78A9.001.0123456-P1-C4/0x21010230/2/25: 32: 0/0/0
lpar29: 29: 568/U78A9.001.0123456-P1-C2/0x21010238/2/29: 32: 0/0/0
lpar5: 5: 521/U78A9.001.0123456-P1-C13/0x21010209/2/5: 32: 0/3/3
lpar5: 5: 520/U78A9.001.0123456-P1-C14/0x21010208/2/5: 32: 0/3/3
lpar9: 9: 529/U78A9.001.0123456-P1-C11/0x21010211/2/9: 32: 0/3/3
lpar9: 9: 528/U78A9.001.0123456-P1-C12/0x21010210/2/9: 32: 0/3/3
cec1: PendingPumpMode=1,CurrentPumpMode=1,OctantCount=8:
OctantID=0,PendingOctCfg=5,CurrentOctCfg=1,PendingMemoryInterleaveMode=2,CurrentMemoryInterleaveMode=2;
OctantID=1,PendingOctCfg=1,CurrentOctCfg=1,PendingMemoryInterleaveMode=2,CurrentMemoryInterleaveMode=2;
OctantID=2,PendingOctCfg=1,CurrentOctCfg=1,PendingMemoryInterleaveMode=2,CurrentMemoryInterleaveMode=2;
OctantID=3,PendingOctCfg=1,CurrentOctCfg=1,PendingMemoryInterleaveMode=2,CurrentMemoryInterleaveMode=2;
OctantID=4,PendingOctCfg=1,CurrentOctCfg=1,PendingMemoryInterleaveMode=2,CurrentMemoryInterleaveMode=2;
OctantID=5,PendingOctCfg=1,CurrentOctCfg=1,PendingMemoryInterleaveMode=2,CurrentMemoryInterleaveMode=2;
OctantID=6,PendingOctCfg=1,CurrentOctCfg=1,PendingMemoryInterleaveMode=2,CurrentMemoryInterleaveMode=2;
OctantID=7,PendingOctCfg=1,CurrentOctCfg=1,PendingMemoryInterleaveMode=2,CurrentMemoryInterleaveMode=2;
Number of BSR arrays: 256,Bytes per BSR array: 4096,Available BSR array: 0;
Available huge page memory(in pages): 0
Configurable huge page memory(in pages): 12
Page Size(in GB): 16
Maximum huge page memory(in pages): 24
Requested huge page memory(in pages): 15
Number of BSR arrays: 256,Bytes per BSR array: 4096,Available BSR array: 0;
Available huge page memory(in pages): 0
Configurable huge page memory(in pages): 12
Page Size(in GB): 16
Maximum huge page memory(in pages): 24
Requested huge page memory(in pages): 15
- To list the virtual machine’s directory entry:
lsvm gpok3
Output is similar to:
gpok3: USER LNX3 PWD 512M 1G G
gpok3: INCLUDE LNXDFLT
gpok3: COMMAND SET VSWITCH VSW2 GRANT LNX3
- For DFM-managed normal power machine, list out the detailed resource information:
lsvm cec
Output is similar to:
cec: HYP Configurable Processors: 16, Avail Processors: 16.
HYP Configurable Memory:32.00 GB(128 regions).
HYP Available Memory: 31.25 GB(125 regions).
HYP Memory Region Size: 0.25 GB(256 MB).
cec: All Physical I/O info:
65535,519,U78AA.001.WZSGVU7-P1-C7,0x21010207,0xffff(Empty Slot)
65535,518,U78AA.001.WZSGVU7-P1-C6,0x21010206,0xffff(Empty Slot)
65535,517,U78AA.001.WZSGVU7-P1-C5,0x21010205,0xffff(Empty Slot)
65535,516,U78AA.001.WZSGVU7-P1-C4,0x21010204,0xffff(Empty Slot)
65535,514,U78AA.001.WZSGVU7-P1-C19,0x21010202,0xffff(Empty Slot)
65535,513,U78AA.001.WZSGVU7-P1-T7,0x21010201,0xc03(USB Controller)
65535,512,U78AA.001.WZSGVU7-P1-T9,0x21010200,0x104(RAID Controller)
cec: Huge Page Memory
Available huge page memory(in pages): 2
Configurable huge page memory(in pages): 2
Page Size(in GB): 16
Maximum huge page memory(in pages): 4
Requested huge page memory(in pages): 2
cec: Barrier Synchronization Register(BSR)
Number of BSR arrays: 256
Bytes per BSR array: 4096
Available BSR array: 256
Note: The lines list in “All Physical I/O info” section represent all the physical I/O resource information. The format is like “owner_lparid,slot_id,physical resource name,drc_index,slot_class_code(class discription)”. The ‘drc index’ is short for Dynamic Resource Configuration Index, it uniquely indicate a physical I/O resource in normal power machine.
For DFM-managed partition on normal power machine, list out the detailed information:
lsvm lpar1
Output is similar to:
lpar1: Lpar Processor Info:
Curr Processor Min: 1.
Curr Processor Req: 16.
Curr Processor Max: 16.
lpar1: Lpar Memory Info:
Curr Memory Min: 0.25 GB(1 regions).
Curr Memory Req: 30.75 GB(123 regions).
Curr Memory Max: 32.00 GB(128 regions).
lpar1: 1,519,U78AA.001.WZSGVU7-P1-C7,0x21010207,0xffff(Empty Slot)
lpar1: 1,518,U78AA.001.WZSGVU7-P1-C6,0x21010206,0xffff(Empty Slot)
lpar1: 1,517,U78AA.001.WZSGVU7-P1-C5,0x21010205,0xffff(Empty Slot)
lpar1: 1,516,U78AA.001.WZSGVU7-P1-C4,0x21010204,0xffff(Empty Slot)
lpar1: 1,514,U78AA.001.WZSGVU7-P1-C19,0x21010202,0xffff(Empty Slot)
lpar1: 1,513,U78AA.001.WZSGVU7-P1-T7,0x21010201,0xc03(USB Controller)
lpar1: 1,512,U78AA.001.WZSGVU7-P1-T9,0x21010200,0x104(RAID Controller)
lpar1: 1/2/2
lpar1: 256.
/opt/xcat/bin/lsvm
mkvm(1)|mkvm.1, chvm(1)|chvm.1, rmvm(1)|rmvm.1
lsxcatd.1¶
lsxcatd - lists xCAT daemon information.
The lsxcat command lists important xCAT daemon (xcatd) information.
-v|--version
Command Version.
-h|--help
Display usage message.
-d|--database
Displays information about the current database being used by xCAT.
-t|--nodetype
Displays whether the node is a Management Node or a Service Node.
-a|--all
Displays all information about the daemon supported by the command.
To display information about the current database:
lsxcatd -d
Output is similar to:
cfgloc=Pg:dbname=xcatdb;host=7.777.47.250|xcatadm dbengine=Pg dbname=xcatdb dbhost=7.777.47.250 dbadmin=xcatadm
To display all information:
lsxcatd -a
Output is similar to:
Version 2.8.5 (git commit 0d4888af5a7a96ed521cb0e32e2c918a9d13d7cc, built Tue Jul 29 02:22:47 EDT 2014) This is a Management Node cfgloc=mysql:dbname=xcatdb;host=9.114.34.44|xcatadmin dbengine=mysql dbname=xcatdb dbhost=9.114.34.44 dbadmin=xcatadmin
/opt/xcat/bin/lsxcatd
mkdef.1¶
mkdef - Use this command to create xCAT data object definitions.
mkdef [-h | --help] [-t object-types]
mkdef [-V | --verbose] [-t object-types] [-o object-names] [-z | --stanza] [-d | --dynamic] [-f | --force] [[-w attr==val] [-w attr=~val] ...] [noderange] [attr=val [attr=val...]]
This command is used to create xCAT object definitions which are stored in the xCAT database. If the definition already exists it will return an error message. The force option may be used to re-create a definition. In this case the old definition will be remove and the new definition will be created.
attr=val [attr=val ...]
Specifies one or more “attribute equals value” pairs, separated by spaces. Attr=val pairs must be specified last on the command line. Use the help option to get a list of valid attributes for each object type.
Note: when creating node object definitions, the ‘groups’ attribute is required.
-d|--dynamic
Use the dynamic option to create dynamic node groups. This option must be used with -w option.
-f|--force
Use the force option to re-create object definitions. This option removes the old definition before creating the new one.
-h|--help
Display usage message.
noderange
A set of comma delimited node names and/or group names. (must be the first parameter) See the “noderange” man page for details on supported formats.
-o object-names
A set of comma delimited object names.
-t object-types
A set of comma delimited object types. Use the help option to get a list of valid object types.
-V|--verbose
Verbose mode.
-w attr==val -w attr=~val ...
Use one or multiple -w flags to specify the selection string that can be used to select objects. The operators ==, !=, =~ and !~ are available. For mkdef commmand, the -w flag only makes sense for creating dynamic node group. Use the help option to get a list of valid attributes for each object type.
- Operator descriptions:
- == Select nodes where the attribute value is exactly this value. != Select nodes where the attribute value is not this specific value. =~ Select nodes where the attribute value matches this regular expression. !~ Select nodes where the attribute value does not match this regular expression.
Note: if the “val” fields includes spaces or any other characters that will be parsed by shell, the “attr<operator>val” needs to be quoted. If the operator is ”!~”, the “attr<operator>val” needs to be quoted using single quote.
-z|--stanza
Indicates that the file being piped to the command is in stanza format. See the xcatstanzafile man page for details on using xCAT stanza files.
-u
Fill in the attributes such as template file, pkglist file and otherpkglist file of osimage object based on the specified parameters. It will search “/install/custom/” directory first, and then “/opt/xcat/share/”. The provmethod and profile must be specified. If osvers or osarch is not specified, the corresponding value of the management node will be used.
Note: this option only works for objtype osimage.
To create a site definition.
mkdef -t site -o clustersite installdir=/xcatinstall
To create a basic node definition.
mkdef -t node -o node01 groups="all,aix"
To re-create the current definition of “node01”.
mkdef -f -t node -o node01 nodetype=osi groups="linux"
(The group definitions are also created if they don’t already exist.)
To create a set of different types of definitions based on information contained in a stanza file.
cat defstanzafile | mkdef -z
To create a group definition called LinuxNodes containing the nodes clstrn01 and clstrn02.
mkdef -t group -o LinuxNodes members="clstrn01,clstrn02"
To create a node definition for an FSP node using the attributes provided by the group fspnodes.
mkdef -t node fspn1 groups=fspnodes nodetype=fsp
To create node definitions for a set of node host names contained in the node range “node1,node2,node3”
mkdef -t node node1,node2,node3 power=hmc groups="all,aix"
To create a dynamic node group definition called HMCMgtNodes containing all the HMC managed nodes”
mkdef -t group -o HMCMgtNodes -d -w mgt==hmc -w cons==hmc
To create a dynamic node group definition called SLESNodes containing all the SLES nodes
mkdef -t group -o SLESNodes -d -w "os=~^sles[0-9]+$"
To create a entry (7.0) in the policy table for user admin1
mkdef -t policy -o 7.0 name=admin1 rule=allow
To create a node definition with nic attributes
mkdef -t node cn1 groups=all nicips.eth0="1.1.1.1|1.2.1.1" nicnetworks.eth0="net1|net2" nictypes.eth0="Ethernet"
To create an osimage definition and fill in attributes automatically.
mkdef redhat6img -u profile=compute provmethod=statelite
$XCATROOT/bin/mkdef
(The XCATROOT environment variable is set when xCAT is installed. The default value is “/opt/xcat”.)
This command is part of the xCAT software product.
chdef(1)|chdef.1, lsdef(1)|lsdef.1, rmdef(1)|rmdef.1, xcatstanzafile(5)|xcatstanzafile.5
mkdsklsnode.1¶
mkdsklsnode - Use this xCAT command to define and initialize AIX/NIM diskless machines.
mkdsklsnode [-h|--help ]
mkdsklsnode [-V|--verbose] [-f|--force] [-n|--newname] [-i osimage_name] [-l location] [-u|--updateSN] [-k|--skipsync] [-p|--primarySN] [-b|--backupSN] [-S|--setuphanfs] noderange [attr=val [attr=val ...]]
This xCAT command can be used to define and/or initialize AIX/NIM diskless machines. Once this step is completed you can use either the xCAT rnetboot command or the rbootseq/rpower commands to initiate a network boot of the nodes.
The command can be used to define and initialize a new NIM machine object or it can be used to reinitialize an existing machine to use a different operating system image.
This command will also create a NIM resolv_conf resource to be used when installing the node. If a resolv_conf resource is not already included in the xCAT osimage definition and if the “domain” and “nameservers” values are set then a new NIM resolv_conf resource will be created and allocated to the nodes.
The “domain” and “nameservers” attributes can be set in either the xCAT “network” definition used by the nodes or in the xCAT cluster “site” definition. The setting in the “network” definition will take priority.
The “search” field of the resolv.conf file will contain a list all the domains listed in the xCAT network definitions and the xCAT site definiton.
The “nameservers” value can either be set to a specific IP address or the “<xcatmaster>” key word. The “<xcatmaster>” key word means that the value of the “xcatmaster” attribute of the node definition will be used in the /etc/resolv.conf file. (I.e. The name of the install server as known by the node.)
You can set the “domain” and “nameservers” attributes by using the chdef command. For example:
chdef -t network -o clstr_net domain=cluster.com nameservers=<xcatmaster>
If the “domain” and “nameservers” attributes are not set in either the nodes “network” definition or the “site” definition then no new NIM resolv_conf resource will be created.
If you are using xCAT service nodes the mkdsklsnode command will automatically determine the correct server(s) for the node and create the NIM definitions on that server(s).
When creating a new NIM machine definition the default is to use the same name as the xCAT node name that is provided.
You can use the “-n” option of the mkdsklsnode command to create and initialize an alternate NIM machine definition for the same physical nodes. This option allows you to set up a new image to use when a node is next rebooted while the node is currently running. This is possible because the NIM name for a machine definition does not have to be the hostname of the node. This allows you to have multiple NIM machine definitions for the same physical node. The naming convention for the new NIM machine name is “<xcat_node_name>_<image_name>”, (Ex. “node01_61spot”). Since all the NIM initialization can be done while the node is running the downtime for for the node is reduced to the time it takes to reboot.
Note: When using the “-n” option make sure that the new osimage you specify and all the NIM resources that are used are different than what are currently being used on the nodes. The NIM resources should not be shared between the old osimage and the new osimage.
You can use the force option to reinitialize a node if it already has resources allocated or it is in the wrong NIM state. This option will reset the NIM node and deallocate resources before reinititializing. Use this option with caution since reinitializing a node will stop the node if it is currently running.
After the mkdsklsnode command completes you can use the lsnim command to check the NIM node definition to see if it is ready for booting the node. (“lsnim -l <nim_node_name>”).
You can supply your own scripts to be run on the management node or on the service node (if their is hierarchy) for a node during the mkdsklsnode command. Such scripts are called prescripts. They should be copied to /install/prescripts dirctory. A table called prescripts is used to specify the scripts and their associated actions. The scripts to be run at the beginning of the mkdsklsnode command are stored in the ‘begin’ column of prescripts table. The scripts to be run at the end of the mkdsklsnode command are stored in the ‘end’ column of prescripts table. Please run ‘tabdump prescripts -d’ command for details. An example for the ‘begin’ or the ‘end’ column is: diskless:myscript1,myscript2. The following two environment variables will be passed to each script: NODES contains all the names of the nodes that need to run the script for and ACTION contains the current current nodeset action, in this case “diskless”. If #xCAT setting:MAX_INSTANCE=number is specified in the script, the script will get invoked for each node in parallel, but no more than number of instances will be invoked at at a time. If it is not specified, the script will be invoked once for all the nodes.
attr=val [attr=val ...]
Specifies one or more “attribute equals value” pairs, separated by spaces. Attr= val pairs must be specified last on the command line. These are used to specify additional values that can be passed to the underlying NIM commands.
Valid values:
duplex
Specifies the duplex setting (optional). Used when defining the NIM machine. Use this setting to configure the client’s network interface. This value can be full or half. The default is full. (ex. “duplex=full”)speed
Specifies the speed setting (optional). Used when defining the NIM machine. This is the communication speed to use when configuring the client’s network interface. This value can be 10, 100, or 1000. The default is 100. (ex. “speed=100”)psize
Specifies the size in Megabytes of the paging space for the diskless node.(optional) Used when initializing the NIM machine. The minimum and default size is 64 MB of paging space. (ex. “psize=256”)sparse_paging
Specifies that the paging file should be created as an AIX sparse file, (ex. “sparse_paging=yes”). The default is “no”.dump_iscsi_port
The tcpip port number to use to communicate dump images from the client to the dump resource server. Normally set by default. This port number is used by a dump resource server.configdump
Specifies the type dump to be collected from the client. The values are “selective”, “full”, and “none”. If the configdump attribute is set to “full” or “selective” the client will automatically be configured to dump to an iSCSI target device. The “selective” memory dump will avoid dumping user data. The “full” memory dump will dump all the memory of the client partition. Selective and full memory dumps will be stored in subdirectory of the dump resource allocated to the client. This attribute is saved in the xCAT osimage definition.
-b |--backupSN
When using backup service nodes only update the backup. The default is to update both the primary and backup service nodes.
-f |--force
Use the force option to reinitialize the NIM machines.
-h |--help
Display usage message.
-i image_name
The name of an existing xCAT osimage definition. If this information is not provided on the command line the code checks the node definition for the value of the “provmethod” attribute. If the “-i” value is provided on the command line then that value will be used to set the “provmethod” attribute of the node definitions.
-k|--skipsync
Use this option to have the mkdsklsnode command skip the NIM sync_roots operation. This option should only be used if you are certain that the shared_root resource does not have to be updated from the SPOT. Normally, when the SPOT is updated, you should do a sync_roots on the shared_root resource.
-l|--location
The directory location to use when creating new NIM resolv_conf resources. The default location is /install/nim.
-n|--newname
Create a new NIM machine object name for the xCAT node. Use the naming convention “<xcat_node_name>_<image_name>” for the new NIM machine definition.
-p|--primarySN
When using backup service nodes only update the primary. The default is to update both the primary and backup service nodes.
-S|--setuphanfs
Setup NFSv4 replication between the primary service nodes and backup service nodes to provide high availability NFS for the compute nodes. This option only exports the /install directory with NFSv4 replication settings, the data synchronization between the primary service nodes and backup service nodes needs to be taken care of through some mechanism.
-u|--updateSN
Use this option if you wish to update the osimages but do not want to define or initialize the NIM client definitions. This option is only valid when the xCAT “site” definition attribute “sharedinstall” is set to either “sns” or “all”.
noderange
A set of comma delimited node names and/or group names. See the “noderange” man page for details on additional supported formats.
-V |--verbose
Verbose mode.
1
Initialize an xCAT node named “node01” as an AIX diskless machine. The xCAT osimage named “61spot” should be used to boot the node.
mkdsklsnode -i 61spot node01
2
Initialize all AIX diskless nodes contained in the xCAT node group called “aixnodes” using the image definitions pointed to by the “provmethod” attribute of the xCAT node definitions.
mkdsklsnode aixnodes
3
Initialize diskless node “clstrn29” using the xCAT osimage called “61dskls”. Also set the paging size to be 128M and specify the paging file be an AIX sparse file.
mkdsklsnode -i 61dskls clstrn29 psize=128 sparse_paging=yes
4
Initialize an xCAT node called “node02” as an AIX diskless node. Create a new NIM machine definition name with the osimage as an extension to the xCAT node name.
mkdsklsnode -n -i 61spot node02
/opt/xcat/bin/mkdsklsnode
This command is part of the xCAT software product.
rmdsklsnode(1)|rmdsklsnode.1
mkflexnode.1¶
mkflexnode - Create a flexible node.
A flexible node is a Partition in a complex. Creating a flexible node is to create a partition which including all the slots defined in the xCAT blade node.
Before creating a flexible node, a general xCAT blade node should be defined. The id attribute of this node should be a node range like ‘a-b’, it means the blades installed in slots ‘a-b’ need to be assigned to the partition. ‘a’ is the start slot, ‘b’ is the end slot. If this partition only have one slot, the slot range can be ‘a’.
The action of creating flexible node will impact the hardware status. Before creating it, the blades in the slot range should be in power off state.
After the creating, use the lsflexnode to check the status of the node.
The noderange only can be a blade node.
1
Create a flexible node base on the xCAT node blade1.
The blade1 should belong to a complex, the id attribute should be set correctly and all the slots should be in power off state.
mkflexnode blade1
/opt/xcat/bin/mkflexnode
lsflexnode(1)|lsflexnode.1, rmflexnode(1)|rmflexnode.1
mkhwconn.1¶
mkhwconn - Sets up connections for CEC and Frame nodes to HMC nodes or hardware server.
mkhwconn [-h| --help]
mkhwconn [-v| --version]
mkhwconn [-V| --verbose] noderange -t [--port port_value]
mkhwconn [-V| --verbose] noderange -s [hmcnode --port port_value]
mkhwconn [-V| --verbose] noderange -p hmc [-P passwd] [--port port_value]
mkhwconn noderange -t [-T tooltype] [--port port_value]
For PPC (with HMC) specific:
This command is used to set up connections for CEC and Frame nodes to HMC nodes. (If the connection already exists, it will not break it.) This command is useful when you have multiple HMCs, each of which will manage a subset of the CECs/Frames. Use mkhwconn to tell each HMC which CECs/Frames it should manage. When using this, you should turn off the self-discovery on each HMC. You also need to put all the HMCs and all the Frames on a single flat service network.
When -t is specified, this command reads the connection information from the xCAT ppc table (e.g. the parent attribute), and read the user/password from the ppcdirect table. Then this command will assign CEC nodes and Frame nodes to HMC nodes.
When -p is specified, this command gets the connection information from command line arguments. If -P is not specified, the default password for CEC and Frame nodes is used.
The flag -s is used to make the connection between the frame and its Service focal point(HMC). Makehwconn will also set the connections between the CECs within this Frame and the HMC. The sfp of the frame/CEC can either be defined in ppc table beforehand or specified in command line after the flag -s. If the user use mkhwconn noderange -s HMC_name, it will not only make the connections but also set the sfp attributes for these nodes in PPC table.
In any case, before running this command, the CEC and Frame nodes need be defined with correct nodetype.nodetype value (cec or frame) and nodehm.mgt value (hmc).
Note: If a CEC belongs to a frame, which has a BPA installed, this CEC should not be assigned to an HMC individually. Instead, the whole frame should be assigned to the HMC.
For PPC (using Direct FSP Management) specific:
It is used to set up connections for CEC and Frame node to Hardware Server on management node (or service node ). It only could be done according to the node definition in xCAT DB. And this command will try to read the user/password from the ppcdirect table first. If fails, then read them from passwd table. Commonly , the username is HMC. If using the ppcdirect table, each CEC/Frame and user/password should be stored in ppcdirect table. If using the passwd table, the key should be “cec” or “frame”, and the related user/password are stored in passwd table.
When --port is specified, this command will create the connections for CECs/Frames whose side in vpd table is equal to port value.
-h|--help
Display usage message.
-t
Read connection information from xCAT DB (ppc and ppcdirect tables). Use this option if you need to connect multiple CECs/Frames to multiple HMCs in a single command.
-p
The HMC node name. Only one HMC nodes can be specified by this flag. To setup connection for multiple HMC nodes, use flag -t.
-P
The password of HMC based CEC/Frame login user(Default user name is ‘HMC’). This flag is optional.
-T
The tooltype is used to communicate to the CEC/Frame. The value could be lpar or fnm. The tooltype value lpar is for xCAT and fnm is for CNM. The default value is “lpar”.
--port
The port value specifies which special side will be used to create the connection to the CEC/Frame. The value could only be specified as “0” or “1” and the default value is “0,1”. If the user wants to use all ports to create the connection, he should not specify this value. If the port value is specified as “0”, in the vpd table, the side column should be A-0 and B-0; If the port value is specified as “1”, the side column should be A-1 and B-1. When making hardware connection between CEC/Frame and HMC, the value is used to specify the fsp/bpa port of the cec/frame and will be organized in order of “A-0,A-1,B-0,B-1”. If any side does not exist, the side would simply be ignored. Generally, only one port of a fsp/bap can be connected while another port be used as backup.
-s
The flag -s is used to make the connection between the frame and its Service Focal Point(HMC). -s flag is not supposed to work with other functional flags.
-V|--verbose
Verbose output.
To setup the connection for all CEC nodes in node group cec to HMC node, according to the definition in xCAT DB:
mkhwconn cec -t
To setup the connection for Frame nodes in node group frame to HMC node hmc1, with password ‘abc123’:
mkhwconn frame -p hmc1 -P abc123
To setup the connections for all CEC nodes in node group cec to hardware server, and the tooltype value is lpar:
mkhwconn cec -t -T lpar
To setup the connections for all cecs nodes in node group cec to hardware server, and the tooltype value is lpar, and the port value is 1:
mkhwconn cec -t -T lpar --port 1
To setup the connection between the frame and it’s SFP node. This command will also set the connections between the CECs within this frame and their SFP node. User need to define HMC_name in the database in advance, but no need to set the sfp attribute for these node, xCAT will set the HMC_name as ppc.sfp for these nodes. The CECs within this frame should have the same sfp attribute as the frame.
mkhwconn cec -s HMC_name -P HMC_passwd
$XCATROOT/bin/mkhwconn
(The XCATROOT environment variable is set when xCAT is installed. The default value is “/opt/xcat”.)
This command is part of the xCAT software product.
lshwconn(1)|lshwconn.1, rmhwconn(1)|rmhwconn.1
mknimimage.1¶
mknimimage - Use this xCAT command to create xCAT osimage definitions and related AIX/NIM resources. The command can also be used to update an existing AIX diskless image(SPOT).
mknimimage [-h | --help ]
mknimimage [-V] -u osimage_name [attr=val [attr=val ...]]
mknimimage [-V] [-f|--force] [-r|--sharedroot] [-D|--mkdumpres] [-l location] [-c|--completeosimage] [-s image_source] [-i current_image] [-p|--cplpp] [-t nimtype] [-m nimmethod] [-n mksysbnode] [-b mksysbfile] osimage_name [attr=val [attr=val ...]]
This command will create both an xCAT osimage definition and the corresponding NIM resource definitions. The command can also be used to update an existing AIX diskless image(SPOT).
The command will also install the NIM master software and configure NIM if needed.
The naming convention for the NIM SPOT resource definition is to use the same name as the xCAT osimage. The naming convention for any other NIM resources that are created is “<osimage_name>_<resource_type>”. (ex. “61image_lpp_source” )
When creating a mksysb image definition you must specify either the “-n” or the “-b” option. The “-n” option can be used to create a mksysb image from an existing NIM client machine. The “-b” option can be used to specify an existing mksysb backup file.
Adding software and configuration files to the osimage.
When creating a diskless osimage definition you also have the option of automatically updating the NIM SPOT resource. You can have additional software installed or you can have configuration files added or updated. To have software installed you must provide either the names of NIM installp_bundle resources or fileset names on the command line using the “attr=val” option. You may also supply the installp flags, RPM flags, emgr flags to use when installing the software.
To have configuration files updated you must provide the full path name of a “synclists” file which contains the the list of actual files to update. The xCAT osimage definition that is created will contain the installp_bundle, otherpkgs, and synclists files that are provided on the command line.
Updating an existing xCAT osimage
If you wish to update an existing diskless image after it has already been created you can use the “-u” (update) option. In this case the xCAT osimage definition will not be updated.
There are two ways to use the update feature.
You can update the osimage definition and run the mknimimage command with no “installp_bundle”, “otherpkgs”, or “synclists” command line values. The information for updating the SPOT will come from the osimage definition only. This has the advantage of keeping a record of any changes that were made to the SPOT.
Or, you could do a more ad hoc update by providing one or more of the “installp_bundle”, “otherpkgs”, or “synclists” values on the command line. If any of these values are provided the mknimimage command will use those values only. The osimage definition will not be used or updated.
WARNING: Installing random RPM packages in a SPOT may have unpredictable consequences. The SPOT is a very restricted environment and some RPM packages may corrupt the SPOT or even hang your management system. Try to be very careful about the packages you install. When installing RPMs, if the mknimimage command hangs or if there are file systems left mounted after the command completes you may need to reboot your management node to recover. This is a limitation of the current AIX support for diskless systems
Copying an xCAT osimage.
You can use the “-i” and “-p” options to copy an existing diskless osimage. To do this you must supply the name of an existing xCAT osimage definition and the name of the new osimage you wish to create. The mknimimage command will do the following:
- create a new xCAT osimage definition using the new name that was specified.
- copy the NIM SPOT resource to a new location and define it to NIM using a new name.
- if the original osimage included a NIM “shared_root” resource then a new shared_root resource will be created for the new SPOT.
- any other resources (or attributes) included in the original osimage will be included in the new osimage definition.
- if the “-p” option is specified then the original NIM lpp_source resource will be copied to a new location and redfined to NIM. (The default would be to use the original lpp_source - to save file system space.)
Additional information
IMPORTANT: The NIM lpp_source and SPOT resources can get quite large. Always make sure that you have sufficient file system space available before running the mknimimage command.
To list the contents of the xCAT osimage definition use the xCAT lsdef command (“lsdef -t osimage -l -o <osimage_name>”).
To check the validity of a SPOT or lpp_source resource
To remove an xCAT osimage definition along with the associated NIM resource definitions use the rmnimimage command. Be careful not to accidently remove NIM resources if they are still needed.
To list a NIM resource definition use the AIX lsnim command (“lsnim -l <resource_name>”).
To check the validity of a SPOT or lpp_source resource use the AIX nim command (“nim -o check <resourec-name>”).
To remove specific NIM resource definitons use the AIX nim command. (“nim -o remove <resource-name>”).
attr=val [attr=val ...]
Specifies one or more “attribute equals value” pairs, separated by spaces. Attr=val pairs must be specified last on the command line.
Currently supported attributes:
bosinst_data
The name of a NIM bosinst_data resource.dump
The name of the NIM dump resource.fb_script
The name of a NIM fb_script resource.home
The name of the NIM home resource.installp_bundle
One or more comma separated NIM installp_bundle resources.lpp_source
The name of the NIM lpp_source resource.mksysb
The name of a NIM mksysb resource.otherpkgs
One or more comma separated installp, emgr, or rpm packages. The packages must have prefixes of ‘I:’, ‘E:’, or ‘R:’, respectively. (ex. R:foo.rpm)paging
The name of the NIM paging resource.resolv_conf
The name of the NIM resolv_conf resource.root
The name of the NIM root resource.script
The name of a NIM script resource.shared_home
The name of the NIM shared_home resource.shared_root
A shared_root resource represents a directory that can be used as a / (root) directory by one or more diskless clients.spot
The name of the NIM SPOT resource.synclists
The fully qualified name of a file containing a list of files to synchronize on the nodes.tmp
The name of the NIM tmp resource.installp_flags
The alternate flags to be passed along to the AIX installp command. (The default for installp_flags is “-abgQXY”.)rpm_flags
The alternate flags to be passed along to the AIX rpm command. (The default for rpm_flags is “-Uvh ”.) The mknimimage command will check each rpm to see if it is installed. It will not be reinstalled unless you specify the appropriate rpm option, such as ‘–replacepkgs’.emgr_flags
The alternate flags to be passed along to the AIX emgr command. (There is no default flags for the emgr command.)dumpsize
The maximum size for a single dump image the dump resource will accept. Space is not allocated until a client starts to dump. The default size is 50GB. The dump resource should be large enough to hold the expected AIX dump and snap data.max_dumps
The maximum number of archived dumps for an individual client. The default is one.snapcollect
Indicates that after a dump is collected then snap data should be collected. The snap data will be collected in the clients dump resource directory. Values are “yes” or “no”. The default is “no”.nfs_vers
Value Specifies the NFS protocol version required for NFS access.nfs_sec
Value Specifies the security method required for NFS access.Note that you may specify multiple “script”, “otherpkgs”, and “installp_bundle” resources by using a comma seperated list. (ex. “script=ascript,bscript”). RPM names may be included in the “otherpkgs” list by using a “R:” prefix(ex. “R:whatever.rpm”). epkg (AIX interim fix package) file names may be included in the “otherpkgs” using the ‘E:’ prefix. (ex. “otherpkgs=E:IZ38930TL0.120304.epkg.Z”).
-b mksysbfile
Used to specify the path name of a mksysb file to use when defining a NIM mksysb resource.
-c|--completeosimage
Complete the creation of the osimage definition passed in on the command line. This option will use any additonal values passed in on the command line and/or it will attempt to create required resources in order to complete the definition of the xCAT osimage. For example, if the osimage definition is missing a spot or shared_root resource the command will create those resources and add them to the osimage definition.
-f|--force
Use the force option to re-create xCAT osimage definition. This option removes the old definition before creating the new one. It does not remove any of the NIM resource definitions named in the osimage definition. Use the rmnimimage command to remove the NIM resources associated with an xCAT osimage definition.
-h |--help
Display usage message.
osimage_name
The name of the xCAT osimage definition. This will be used as the name of the xCAT osimage definition as well as the name of the NIM SPOT resource.
-D|--mkdumpres
Create a diskless dump resource.
-i current_image
The name of an existing xCAT osimage that should be copied to make a new xCAT osimage definition. Only valid when defining a “diskless” or “dataless” type image.
-l location
The directory location to use when creating new NIM resources. The default location is /install/nim.
-m nimmethod
Used to specify the NIM installation method to use. The possible values are “rte” and “mksysb”. The default is “rte”.
-n mksysbnode
The xCAT node to use to create a mksysb image. The node must be a defined as a NIM client machine.
-p|--cplpp
Use this option when copying existing diskless osimages to indicate that you also wish to have the lpp_resource copied. This option is only valid when using the “-i” option.
-r|--sharedroot
Use this option to specify that a NIM “shared_root” resource be created for the AIX diskless nodes. The default is to create a NIM “root” resource. This feature is only available when using AIX version 6.1.4 or beyond. See the AIX/NIM documentation for a description of the “root” and “shared_root” resources.
-s image_source
The source of software to use when creating the new NIM lpp_source resource. This could be a source directory or a previously defined NIM lpp_source resource name.
-t nimtype
Used to specify the NIM machine type. The possible values are “standalone”, “diskless” or “dataless”. The default is “standalone”.
-u
Used to update an AIX/NIM SPOT resource with additional software and configuration files. This option is only valid for xCAT diskless osimage objects. The SPOT resource associated with the xCAT osimage definition will be updated. This option can also be used to update the nfs_vers attribute from NFSv3 to NFSv4 for the NIM resources associated with diskful or diskless image.
-V |--verbose
Verbose mode.
- Create an osimage definition and the basic NIM resources needed to do a NIM “standalone” “rte” installation of node “node01”. Assume the software contained on the AIX product media has been copied to the /AIX/instimages directory.
mknimimage -s /AIX/instimages 61image
- Create an osimage definition that includes some additional NIM resources.
mknimimage -s /AIX/instimages 61image installp_bundle=mybndlres,addswbnd
This command will create lpp_source, spot, and bosinst_data resources using the source specified by the “-s” option. The installp_bundle information will also be included in the osimage definition. The mybndlres and addswbnd resources must be created before using this osimage definition to install a node.
- Create an osimage definition that includes a mksysb image and related resources.
mknimimage -m mksysb -n node27 newsysb spot=myspot bosinst_data=mybdata
This command will use node27 to create a mksysb backup image and use that to define a NIM mksysb resource. The osimage definition will contain the name of the mksysb resource as well as the spot and bosinst_data resource.
- Create an osimage definition using a mksysb image provided on the command line.
mknimimage -m mksysb -b /tmp/backups/mysysbimage newsysb spot=myspot bosinst_data=mybdata
This command defines a NIM mksysb resource using mysysbimage.
- Create an osimage definition and create the required spot definition using the mksysb backup file provided on the command line.
mknimimage -m mksysb -b /tmp/backups/mysysbimage newsysb bosinst_data=mybdata
This command defines a NIM mksysb resource and a spot definition using mysysbimage.
- Create a diskless image called 61dskls using the AIX source files provided in the /AIX/instimages directory.
mknimimage -t diskless -s /AIX/instimages 61dskls
- Create a diskless image called “614dskls” that includes a NIM “shared_root” and a “dump” resource. Use the existing NIM lpp_resource called “614_lpp_source”. Also specify verbose output.
mknimimage -V -r -D -t diskless -s 614_lpp_source 614dskls snapcollect=yes
The “snapcollect” attribute specifies that AIX “snap” data should be include when a system dump is initiated.
- Create a new diskless image by copying an existing image.
mknimimage -t diskless -i 61cosi 61cosi_updt1
Note: If you also wish to have the original lpp_source copied and defined use the -p option.
mknimimage -t diskless -i 61cosi -p 61cosi_updt1
- Create a diskless image using an existing lpp_source resource named “61cosi_lpp_source” and include NIM tmp and home resources. This assumes that the “mytmp” and “myhome” NIM resources have already been created by using NIM commands.
mknimimage -t diskless -s 61cosi_lpp_source 611cosi tmp=mytmp home=myhome
- Create a diskless image and update it with additional software using rpm flags and configuration files.
mknimimage -t diskless -s 61cosi_lpp_source 61dskls otherpkgs=I:fset1,R:foo.rpm,E:IZ38930TL0.120304.epkg.Z synclists=/install/mysyncfile rpm_flags=”-i --nodeps”
The xCAT osimage definition created by this command will include the “otherpkgs” and “synclists” values. The NIM SPOT resource associated with this osimage will be updated with the additional software using rpm flags “-i –nodeps” and configuration files.
- Update an existing diskless image (AIX/NIM SPOT) using the information saved in the xCAT “61dskls” osimage definition. Also specify verbose messages.
mknimimage -V -u 61dskls
- Update an existing diskless image called “61dskls”. Install the additional software specified in the NIM “bndres1” and “bndres2” installp_bundle resources using the installp flags “-agcQX”. (The NIM “bndres1” and “bndres2” definitions must be created before using them in this command.)
mknimimage -u 61dskls installp_bundle=bndres1,bndres2 installp_flags=”-agcQX”
Note that when “installp_bundle”, “otherpkgs”, or “synclists” values are specified with the “-u” option then the xCAT osimage definiton is not used or updated.
- Update an existing image to support NFSv4. Also specify verbose messages.
mknimimage -V -u 61dskls nfs_vers=4
/opt/xcat/bin/mknimimage
This command is part of the xCAT software product.
rmnimimage(1)|rmnimimage.1
mkvlan.1¶
mkvlan - It takes a list of nodes and create a private tagged vlan for them.
mkvlan [vlanid] -n | --nodes noderange [-t | --net subnet] [-m | --mask netmask] [-p | --prefix hostname_prefix] [-i | --interface nic]
mkvlan [-h | --help]
mkvlan [-v | --version]
The mkvlan command takes a list of nodes and move them to a private vlan.
This command will configure the switch to create a new tagged vlan on the given nic. The primary nic will be used if the nic is not specified. The new vlan ID is given by the command. However, if it is omitted, xCAT will automatically generate the new vlan ID by querying all the switches involved and finding out the smallest common number that is not used by any existing vlans. The subnet and the netmask for the vlan will be derived from the value of “vlannets” and “vlanmasks” from the site table if -t and -m are not specified. The following are the default site table entires:
vlannets="|(\d+)|10.($1+0).0.0|";
vlanmask="255.255.0.0";
The vlan network will be entered in the networks table. The nodes will be added to the vlan using the vlan tagging technique. And the new IP addresses and new hostnames will be assigned to the nodes. The -p flag specifies the node hostname prefix for the nodes. If it is not specified, by default, the hostnames for the nodes are having the following format:
v<vlanid>nY where Y is the node number. For example, the hostname for node 5 on vlan 10 is v10n5.
The switch.vlan will be updated with the new vlan id for the node for standaline nodes. For KVM guests, the vm.nics identifies which vlan this node belongs to. For example: vl3 means this node is in vlan 3.
If there are more than one switches involved in the vlan, the ports that connect to the switches need to entered in switches.linkports with the following format:
<port numner>:switch,<port number>:switch....
For example:
"42:switch1,43:switch2"
This command will automatically configure the cross-over ports if the given nodes are on different switches.
For added security, the root guard and bpdu guard will be enabled for the ports in this vlan. However, the guards will not be disabled if the ports are removed from the vlan using chvlan or rmvlan commands. To disable them, you need to use the switch command line interface. Please refer to the switch command line interface manual to see how to disable the root guard and bpdu guard for a port.
vlanid is a unique vlan number. If it is omitted, xCAT will automatically generate the new vlan ID by querying all the switches involved and finding out the smallest common number that is not used by any existing vlans. Use lsvlan to find out the existing vlan ids used by xCAT.
-n|--nodes The nodes or groups to be included in the vlan. It can be stand alone nodes or KVM guests. It takes the noderange format. Please check the man page for noderange for details.
-t|--net The subnet for the vlan.
-m|--mask The netmask for the vlan
-p|--prefix The prefix the the new hostnames for the nodes in the vlan.
-i|--interface The interface name where the vlan will be tagged on. If omitted, the xCAT management network will be assumed. For FVM, this is the interface name on the host.
-h|--help Display usage message.
-v|--version The Command Version.
To start, the xCAT switches and switches table needs to be filled with switch and port info for the nodes. For example, the swith table will look like this:
#node,switch,port,vlan,interface,comments,disable “node1”,”switch1”,”10”,,,, “node1”,”switch2”,”1”,,”eth1”,, “node2”,”switch1”,”11”,,”primary”,, “node2”,”switch2”,”2”,,”eth1”,, “node3”,”switch1”,”12”,,”primary:eth0”,, “node3”,”switch2”,”3”,,”eth1”,,
Please note that the interface value for the management (primary) network can be empty, the word “primary” or “primary:ethx”. For other networks, the interface attribute must be specified.
The following is an example of the switches table
#switch,snmpversion,username,password,privacy,auth,linkports,sshusername,sshpassword,switchtype,comments,disable “switch1”,”3”,”username”,”passw0rd”,,”sha”,”48:switch2”,,,,, “switch2”,”2”,,,,,”43:switch1”,,,,,
To make a private vlan for node1, node2 and node3
mkvlan -n node1,node2,node3
The vlan will be created on eth0 for the nodes.
To make a private vlan for node1, node2 and node3 on eth1,
mkvlan -n node1,node2,node3 -i eth1
TO make a private vlan for node1, node2 with given subnet and netmask.
mkvlan -n node1,node2,node3 -t 10.3.2.0 -m 255.255.255.0
To make a private vlan for KVM guests node1 and node2
chtab key=usexhrm site.vlaue=1 mkdef node1 arch=x86_64 groups=kvm,all installnic=mac primarynic=mac mgt=kvm netboot=pxe nfsserver=10.1.0.204 os=rhels6 profile=compute provmethod=install serialport=0 serialspeed=115200 vmcpus=1 vmhost=x3650n01 vmmemory=512 vmnics=br0 vmstorage=nfs://10.1.0.203/vms mkdef node2 arch=x86_64 groups=kvm,all installnic=mac primarynic=mac mgt=kvm netboot=pxe nfsserver=10.1.0.204 os=rhels6 profile=compute provmethod=install serialport=0 serialspeed=115200 vmcpus=1 vmhost=x3650n01 vmmemory=512 vmnics=br0 vmstorage=nfs://10.1.0.203/vms mkvlan -n node1,node2 mkvm node1,node2 -s 20G rpower node1,node2 on rinstall node1,node2
/opt/xcat/bin/mkvlan
chvlan(1)|chvlan.1, rmvlan(1)|rmvlan.1, lsvlan(1)|lsvlan.1
mkvm.1¶
mkvm - Creates HMC-, DFM-, IVM-, and zVM-managed partitions or other virtual machines.
mkvm [-V| --verbose] noderange -i id -l singlenode
mkvm [-V| --verbose] noderange -c destcec -p profile
mkvm [-V| --verbose] noderange --full
mkvm noderange [--full]
- mkvm noderange [vmcpus=min/req/max] [vmmemory=min/req/max]
- [vmphyslots=drc_index1,drc_index2...] [vmothersetting=hugepage:N,bsr:N] [vmnics=vlan1[,vlan2..]] [vmstorage=<N|viosnode:slotid>] [--vios]
mkvm noderange [-m|--master mastername] [-s|--size disksize] [--mem memsize] [--cpus cpucount] [-f|--force]
mkvm noderange [-s | --size disksize] [--mem memsize] [--cpus cpucount]
mkvm noderange [directory_entry_file_path]
mkvm noderange [source_virtual_machine] [pool= disk_pool]
The first form of mkvm command creates new partition(s) with the same profile/resources as the partition specified by singlenode. The -i and noderange specify the starting numeric partition number and the noderange for the newly created partitions, respectively. The LHEA port numbers and the HCA index numbers will be automatically increased if they are defined in the source partition.
The second form of this command duplicates all the partitions from the source specified by profile to the destination specified by destcec. The source and destination CECs can be managed by different HMCs.
Please make sure the nodes in the noderange is defined in the nodelist table and the mgt is set to ‘hmc’ in the nodehm table before running this command.
Please note that the mkvm command currently only supports creating standard LPARs, not virtual LPARs working with VIOS server.
With option full, a partition using all the resources on a normal power machine will be created.
If no option is specified, a partition using the parameters specified with attributes such as ‘vmcpus’, ‘vmmory’, ‘vmphyslots’, ‘vmothersetting’, ‘vmnics’, ‘vmstorage’ will be created. Those attributes can either be specified with ‘*def’ commands running before or be specified with this command.
The mkvm command creates new virtual machine(s) with the disksize size of hard disk, memsize size of memory and cpucount number of cpu.
For KVM: If -f | --force is specified, the storage will be destroyed first if it existed.
The first form of mkvm creates a new virtual machine based on a directory entry.
The second form of this creates a new virtual machine with the same profile/resources as the specified node (cloning).
-h|--help
Display usage message.
-c
The cec (fsp) name for the destination.
--cpus
The cpu count which will be created for the kvm/vmware virtual machine.
--full
Request to create a new full system partition for each CEC.
vmcpus=value vmmemory=value vmphyslots=value vmothersetting=value vmnics=value vmstorage=value [--vios]
To specify the parameters which are used to create a partition. The vmcpus, vmmemory are necessay, and the value specified with this command have a more high priority. If the value of any of the three options is not specified, the corresponding value specified for the node object will be used. If any of the three attributes is neither specified with this command nor specified with the node object, error information will be returned. To reference to lsvm(1)|lsvm.1 for more information about ‘drc_index’ for vmphyslots.
The option vios is used to specify the partition that will be created is a VIOS partition. If specified, the value for vmstorage shall be number which indicate the number of vSCSI server adapter will be created, and if no value specified for vmphyslots, all the physical slot of the power machine will be asigned to VIOS partition. If not specified, it shall be in form of vios_name:server_slotid to specify the vios and the virtual slot id of the vSCSI server adapter that will be connected from the Logical partition.
-f|--force
If -f|--force is specified, the storage will be destroyed first if it existed.
-i
Starting numeric id of the newly created partitions.
-l
The partition name of the source.
--mem
The memory size which will be used for the new created kvm/vmware virtual machine. Unit is Megabyte.
-p
The file that contains the profiles for the source partitions.
-s|--size
The size of storage which will be created for the kvm/vmware virtual machine.
-v|--version
Command Version.
-V|--verbose
Verbose output.
- To create a new HMC-managed partition lpar5 based on the profile/resources of lpar4, enter:
mkdef -t node -o lpar5 mgt=hmc groups=all
then:
mkvm lpar5 -i 5 -l lpar4
Output is similar to:
lpar5: Success
- To create new HMC-managed partitions lpar5-lpar8 based on the profile/resources of lpar4, enter:
mkdef -t node -o lpar5-lpar8 mgt=hmc groups=all
then:
mkvm lpar5-lpar8 -i 5 -l lpar4
Output is similar to:
lpar5: Success
lpar6: Success
lpar7: Success
lpar8: Success
- To duplicate all the HMC-managed partitions associated with cec01 on cec02, first save the lpars from cec01 to a file:
lsvm lpar01-lpar04 > /tmp/myprofile
then create lpars on cec02:
mkvm lpar05-lpar08 -c cec02 -p /tmp/myprofile
Output is similar to:
lpar5: Success
lpar6: Success
lpar7: Success
lpar8: Success
- To duplicate all the HMC-managed partitions associated with cec01 on cec02, one is for cec01, the other is for cec02:
mkdef -t node -o lpar5,lpar6 mgt=hmc groups=all
chtab node=lpar5 ppc.parent=cec01
chtab node=lpar6 ppc.parent=cec02
then create lpars on cec01 and cec02:
mkvm lpar5,lpar6 --full
Output is similar to:
lpar5: Success
lpar6: Success
- To create a new zVM virtual machine (gpok3) based on a directory entry:
mkvm gpok3 /tmp/dirEntry.txt
Output is similar to:
gpok3: Creating user directory entry for LNX3... Done
- To clone a new zVM virtual machine with the same profile/resources as the specified node:
mkvm gpok4 gpok3 pool=POOL1
Output is similar to:
gpok4: Cloning gpok3
gpok4: Linking source disk (0100) as (1100)
gpok4: Linking source disk (0101) as (1101)
gpok4: Stopping LNX3... Done
gpok4: Creating user directory entry
gpok4: Granting VSwitch (VSW1) access for gpok3
gpok4: Granting VSwitch (VSW2) access for gpok3
gpok4: Adding minidisk (0100)
gpok4: Adding minidisk (0101)
gpok4: Disks added (2). Disks in user entry (2)
gpok4: Linking target disk (0100) as (2100)
gpok4: Copying source disk (1100) to target disk (2100) using FLASHCOPY
gpok4: Mounting /dev/dasdg1 to /mnt/LNX3
gpok4: Setting network configuration
gpok4: Linking target disk (0101) as (2101)
gpok4: Copying source disk (1101) to target disk (2101) using FLASHCOPY
gpok4: Powering on
gpok4: Detatching source disk (0101) at (1101)
gpok4: Detatching source disk (0100) at (1100)
gpok4: Starting LNX3... Done
- To create a new kvm/vmware virtual machine with 10G storage, 2048M memory and 2 cpus.
mkvm vm1 -s 10G --mem 2048 --cpus 2
- To create a full partition on normal power machine.
First, define a node object:
mkdef -t node -o lpar1 mgt=fsp cons=fsp nodetype=ppc,osi id=1 hcp=cec parent=cec hwtype=lpar groups=lpar,all
Then, create the partion on the specified cec.
mkvm lpar1 --full
The output is similar to:
lpar1: Done
To query the resources allocated to node ‘lpar1’
lsvm lpar1
The output is similar to:
lpar1: Lpar Processor Info:
Curr Processor Min: 1.
Curr Processor Req: 16.
Curr Processor Max: 16.
lpar1: Lpar Memory Info:
Curr Memory Min: 0.25 GB(1 regions).
Curr Memory Req: 30.75 GB(123 regions).
Curr Memory Max: 32.00 GB(128 regions).
lpar1: 1,519,U78AA.001.WZSGVU7-P1-C7,0x21010207,0xffff(Empty Slot)
lpar1: 1,518,U78AA.001.WZSGVU7-P1-C6,0x21010206,0xffff(Empty Slot)
lpar1: 1,517,U78AA.001.WZSGVU7-P1-C5,0x21010205,0xffff(Empty Slot)
lpar1: 1,516,U78AA.001.WZSGVU7-P1-C4,0x21010204,0xffff(Empty Slot)
lpar1: 1,514,U78AA.001.WZSGVU7-P1-C19,0x21010202,0xffff(Empty Slot)
lpar1: 1,513,U78AA.001.WZSGVU7-P1-T7,0x21010201,0xc03(USB Controller)
lpar1: 1,512,U78AA.001.WZSGVU7-P1-T9,0x21010200,0x104(RAID Controller)
lpar1: 1/2/2
lpar1: 256.
Note: The ‘parent’ attribute for node ‘lpar1’ is the object name of physical power machine that the full partition will be created on.
- To create a partition using some of the resources on normal power machine.
Option 1:
After a node object is defined, the resources that will be used for the partition shall be specified like this:
chdef lpar1 vmcpus=1/4/16 vmmemory=1G/4G/32G vmphyslots=0x21010201,0x21010200 vmothersetting=bsr:128,hugepage:2
Then, create the partion on the specified cec.
mkvm lpar1
Option 2:
mkvm lpar1 vmcpus=1/4/16 vmmemory=1G/4G/32G vmphyslots=0x21010201,0x21010200 vmothersetting=bsr:128,hugepage:2
The outout is similar to:
lpar1: Done
Note: The ‘vmplyslots’ specify the drc index of the physical slot device. Every drc index shall be delimited with ‘,’. The ‘vmothersetting’ specify two kinds of resource, bsr(Barrier Synchronization Register) specified the num of BSR arrays, hugepage(Huge Page Memory) specified the num of huge pages.
To query the resources allocated to node ‘lpar1’
lsvm lpar1
The output is similar to:
lpar1: Lpar Processor Info:
Curr Processor Min: 1.
Curr Processor Req: 4.
Curr Processor Max: 16.
lpar1: Lpar Memory Info:
Curr Memory Min: 1.00 GB(4 regions).
Curr Memory Req: 4.00 GB(16 regions).
Curr Memory Max: 32.00 GB(128 regions).
lpar1: 1,513,U78AA.001.WZSGVU7-P1-T7,0x21010201,0xc03(USB Controller)
lpar1: 1,512,U78AA.001.WZSGVU7-P1-T9,0x21010200,0x104(RAID Controller)
lpar1: 1/2/2
lpar1: 128.
- To create a vios partition using some of the resources on normal power machine.
mkvm viosnode vmcpus=1/4/16 vmmemory=1G/4G/32G vmphyslots=0x21010201,0x21010200 vmnics=vlan1 vmstorage=5 --vios
The resouces for the node is similar to:
viosnode: Lpar Processor Info:
Curr Processor Min: 1.
Curr Processor Req: 4.
Curr Processor Max: 16.
viosnode: Lpar Memory Info:
Curr Memory Min: 1.00 GB(4 regions).
Curr Memory Req: 4.00 GB(16 regions).
Curr Memory Max: 32.00 GB(128 regions).
viosnode: 1,513,U78AA.001.WZSGVU7-P1-T7,0x21010201,0xc03(USB Controller)
viosnode: 1,512,U78AA.001.WZSGVU7-P1-T9,0x21010200,0x104(RAID Controller)
viosnode: 1,0,U8205.E6B.0612BAR-V1-C,0x30000000,vSerial Server
viosnode: 1,1,U8205.E6B.0612BAR-V1-C1,0x30000001,vSerial Server
viosnode: 1,3,U8205.E6B.0612BAR-V1-C3,0x30000003,vEth (port_vlanid=1,mac_addr=4211509276a7)
viosnode: 1,5,U8205.E6B.0612BAR-V1-C5,0x30000005,vSCSI Server
viosnode: 1,6,U8205.E6B.0612BAR-V1-C6,0x30000006,vSCSI Server
viosnode: 1,7,U8205.E6B.0612BAR-V1-C7,0x30000007,vSCSI Server
viosnode: 1,8,U8205.E6B.0612BAR-V1-C8,0x30000008,vSCSI Server
viosnode: 1,9,U8205.E6B.0612BAR-V1-C9,0x30000009,vSCSI Server
viosnode: 0/0/0
viosnode: 0.
/opt/xcat/bin/mkvm
chvm(1)|chvm.1, lsvm(1)|lsvm.1, rmvm(1)|rmvm.1
mkzone.1¶
mkzone - Defines a new zone in the cluster.
mkzone <zonename> [--defaultzone] [-k full path to the ssh RSA private key] [-a noderange] [-g] [-f] [-s yes|no] [-V]
mkzone [-h | -v]
The mkzone command is designed to divide the xCAT cluster into multiple zones. The nodes in each zone will share common root ssh keys. This allows the nodes in a zone to be able to as root ssh to each other without password, but cannot do the same to any node in another zone. All zones share a common xCAT Management Node and database including the site table, which defines the attributes of the entire cluster. The mkzone command is only supported on Linux ( No AIX support). The nodes are not updated with the new root ssh keys by mkzone. You must run updatenode -k or xdsh -K to the nodes to update the root ssh keys to the new generated zone keys. This will also sync any service nodes with the zone keys, if you have a hierarchical cluster. Note: if any zones in the zone table, there must be one and only one defaultzone. Otherwise, errors will occur.
-h | --help
Displays usage information.
-v | --version
Displays command version and build date.
-k | --sshkeypath full path to the ssh RSA private key
This is the path to the id_rsa key that will be used to build root’s ssh keys for the zone. If -k is used, it will generate the ssh public key from the input ssh RSA private key and store both in /etc/xcat/sshkeys/<zonename>/.ssh directory. If -f is not used, then it will generate a set of root ssh keys for the zone and store them in /etc/xcat/sshkeys/<zonename>/.ssh.
--default
if –defaultzone is input, then it will set the zone defaultzone attribute to yes; otherwise it will set to no. if –defaultzone is input and another zone is currently the default, then the -f flag must be used to force a change to the new defaultzone. If -f flag is not use an error will be returned and no change made. Note: if any zones in the zone table, there must be one and only one defaultzone. Otherwise, errors will occur.
-a | --addnoderange noderange
For each node in the noderange, it will set the zonename attribute for that node to the input zonename. If the -g flag is also on the command, then it will add the group name “zonename” to each node in the noderange.
-s| --sshbetweennodes yes|no
If -s entered, the zone sshbetweennodes attribute will be set to yes or no. It defaults to yes. When this is set to yes, then ssh will be setup to allow passwordless root access between nodes. If no, then root will be prompted for a password when running ssh between the nodes in the zone.
-f | --force
Used with the (–defaultzone) flag to override the current default zone.
-g | --assigngroup
Used with the (-a) flag to create the group zonename for all nodes in the input noderange.
-V | --Verbose
Verbose mode.
*
To make a new zone1 using defaults , enter:
mkzone zone1
- Note: with the first mkzone, you will automatically get the xcatdefault zone created as the default zone. This zone uses ssh keys from
- <roothome>/.ssh directory.
*
*
To make a new zone2A using the ssh id_rsa private key in /root/.ssh:
mkzone zone2A -k /root/.ssh
*
To make a new zone3 and assign the noderange compute3 to the zone enter:
mkzone zone3 -a compute3
*
To make a new zone4 and assign the noderange compute4 to the zone and add zone4 as a group to each node enter:
mkzone zone4 -a compute4 -g
*
To make a new zone5 and assign the noderange compute5 to the zone and add zone5 as a group to each node but not allow passwordless ssh between the nodes enter:
mkzone zone5 -a compute5 -g -s no
Files
/opt/xcat/bin/mkzone/
Location of the mkzone command.
chzone(1)|chzone.1, rmzone(1)|rmzone.1, xdsh(1)|xdsh.1, updatenode(1)|updatenode.1
monadd.1¶
monadd - Registers a monitoring plug-in to the xCAT cluster.
monadd [-h| –help]
monadd [-v| –version]
monadd name [-n|–nodestatmon] [-s|–settings settings]
This command is used to register a monitoring plug-in module to monitor the xCAT cluster. The plug-in module will be added to the xCAT monitoring database table and the configuration scripts for the monitoring plug-in, if any, will be added to the postscripts table. A monitoring plug-in module acts as a bridge that connects a 3rd party monitoring software and the xCAT cluster. A configuration script is used to configure the 3rd party software. Once added to the <postscripts> table, it will be invoked on the nodes during node deployment stage.
name is the name of the monitoring plug-in module. For example, if the the name is called xxx, then the actual file name that the xcatd looks for is /opt/xcat/lib/perl/xCAT_monitoring/xxx.pm. Use monls -a command to list all the monitoring plug-in modules that can be used.
settings is the monitoring plug-in specific settings. It is used to customize the behavior of the plug-in or configure the 3rd party software. Format: -s key-value -s key=value ... Please note that the square brackets are needed here. Use monls name -d command to look for the possbile setting keys for a plug-in module.
-h | --help
Display usage message.
-n | --nodestatmon
Indicate that this monitoring plug-in will be used for feeding the node liveness status to the xCAT nodelist table.
-s | --settings
Specifies the plug-in specific settings. These settings will be used by the plug-in to customize certain entities for the plug-in or the third party monitoring software. e.g. -s mon_interval=10 -s toggle=1.
Command Version.
To register gangliamon plug-in module (which interacts with Ganglia monitoring software) to monitor the xCAT cluster, enter:
monadd gangliamon
To register rmcmon plug-in module (which interacts with IBM’s RSCT monitoring software) to monitor the xCAT cluster and have it feed the node liveness status to xCAT’s nodelist table, enter:
monadd rmcmon -n
This will also add the configrmcnode to the postscripts table. To view the content of the postscripts table, enter:
tabdump postscritps #node,postscripts,comments,disable "service","servicenode",, "xcatdefaults","syslog,remoteshell,configrmcnode",,
To register xcatmon plug-in module to feed the node liveness status to xCAT’s nodelist table, enter:
monadd xcatmon -n -s ping-interval=2
where 2 is the number of minutes between the pings.
/opt/xcat/bin/monadd
monls(1)|monls.1, monrm(1)|monrm.1, monstart(1)|monstart.1, monstop(1)|monstop.1, moncfg(1)|moncfg.1, mondecfg(1)|mondecfg.1
moncfg.1¶
moncfg - Configures a 3rd party monitoring software to monitor the xCAT cluster.
This command is used to configure a 3rd party monitoring software to monitor the xCAT cluster. For example, it modifies the configration file for the monitoring software so that the nodes can be included in the monitoring domain. The operation is performed on the management node and the service nodes of the given nodes. The operation will also be performed on the nodes if the -r option is specified, though the configuration of the nodes is usually performed during the node deployment stage.
name is the name of the monitoring plug-in module. For example, if the the name is called xxx, then the actual file name that the xcatd looks for is /opt/xcat/lib/perl/xCAT_monitoring/xxx.pm. Use monls -a command to list all the monitoring plug-in modules that can be used.
noderange specifies the nodes to be monitored. If omitted, all nodes will be monitored.
-h | --help Display usage message.
-r | --remote Specifies that the operation will also be performed on the nodes.
- To configure the management node and the service nodes for ganglia monitoring, enter:
moncfg gangliamon
- To configure the management node, nodes and their service nodes for ganglia monitoring, enter:
moncfg gangliamon -r
/opt/xcat/bin/moncfg
monls(1)|monls.1, mondecfg(1)|mondecfg.1, monadd(1)|monadd.1, monrm(1)|monrm.1, monstart(1)|monstart.1, monstop(1)|monstop.1
mondecfg.1¶
mondecfg - Deconfigures a 3rd party monitoring software from monitoring the xCAT cluster.
This command is used to deconfigure a 3rd party monitoring software from monitoring the xCAT cluster. The operation is performed on the management node and the service nodes of the given nodes. The operation will also be performed on the nodes if the -r option is specified. The deconfigration operation will remove the nodes from the 3rd party software’s monitoring domain.
name is the name of the monitoring plug-in module. Use monls command to list all the monitoring plug-in modules that can be used.
noderange specified the nodes to be deconfigured. If omitted, all nodes will be deconfigured.
-h | --help Display usage message.
-r | --remote Specifies that the operation will also be performed on the nodes.
- To deconfigure the management node and the service nodes from the ganglia monitoring, enter:
mondecfg gangliamon
- To deconfigure the management node, nodes and their service nodes from the ganglia monitoring, enter:
mondecfg gangliamon -r
/opt/xcat/bin/mondecfg
monls(1)|monls.1, moncfg(1)|moncfg.1, monadd(1)|monadd.1, monrm(1)|monrm.1, monstart(1)|monstart.1, monstop(1)|monstop.1
monls.1¶
monls - Lists monitoring plug-in modules that can be used to monitor the xCAT cluster.
monls [-h| –help]
monls [-v| –version]
monls *name* [-d|–description]
monls [-a|–all] [-d|–description]
This command is used to list the status, desctiption, the configuration scripts and the settings of one or all of the monitoring plug-in modules.
name is the name of the monitoring plug-in module.
-a | --all Searches the XCATROOT/lib/perl/xCAT_monitoring directory and reports all the monitoring plug-in modules. If nothing is specified, the list is read from the monitoring tabel.
-d | --description Display the description of the plug-in modules. The description ususally contains the possible settings.
-h | --help Display usage message.
- To list the status of all the monitoring plug-in modules from the monitoring table, enter:
monls
The output looks like this:
xcatmon monitored node-status-monitored
snmpmon not-monitored
- To list the status of all the monitoring plug-in modules including the ones that are not in the monitoring table, enter
monls -a
The output looks like this:
xcatmon monitored node-status-monitored
snmpmon not-monitored
gangliamon not-monitored
rmcmon monitored
nagiosmon not-monitored
- To list the status and the desciption for snmpmon module, enter:
monls snmpmon -d
/opt/xcat/bin/monls
monadd(1)|monadd.1, monrm(1)|monrm.1, monstart(1)|monstart.1, monstop(1)|monstop.1, moncfg(1)|moncfg.1, mondecfg(1)|mondecfg.1
monrm.1¶
monrm - Unregisters a monitoring plug-in module from the xCAT cluster.
This command is used to unregister a monitoring plug-in module from the monitoring table. It also removes any configuration scripts associated with the monitoring plug-in from the postscripts table. A monitoring plug-in module acts as a bridge that connects a 3rd party monitoring software and the xCAT cluster. A configuration script is used to configure the 3rd party software. Once added to the postscripts table, it will be invoked on the nodes during node deployment stage.
name is the name of the monitoring plug-in module in the monitoring table. Use monls command to list all the monitoring plug-in modules that can be used.
1.To unregister gangliamon plug-in module (which interacts with Ganglia monitoring software) from the xCAT cluster, enter:
monrm gangliamon
Please note that gangliamon must have been registered in the xCAT monitoring table. For a list of registered plug-in modules, use command monls.
/opt/xcat/bin/monrm
monls(1)|monls.1, monadd(1)|monadd.1, monstart(1)|monstart.1, monstop(1)|monstop.1, moncfg(1)|moncfg.1, mondecfg(1)|mondecfg.1
monshow.1¶
monshow - Shows event data for monitoring.
monshow [-h| –help]
monshow [-v| –version]
monshow name [noderange] [-s] [-t time] [-a attributes] [-w attr<operatorval [-w attr<operator>val] ... ][-o {p|e}]>
This command displays the events that happened on the given nodes or the monitoring data that is collected from the given nodes for a monitoring plugin.
name is the name of the monitoring plug-in module to be invoked.
noderange is a list of nodes to be showed for. If omitted, the data for all the nodes will be displayed.
-h | --help Display usage message.
-v | -**-version ** Command Version.
-s shows the summary data.
-t specifies a range of time for the data, The default is last 60 minutes. For example -t 6-4, it will display the data from last 6 minutes to 4 minutes; If it is -t 6, it will display the data from last 6 minutes until now.
-a specifies a comma-separated list of attributes or metrics names. The default is all.
-w specify one or multiple selection string that can be used to select events. The operators ==, !=, =,!,>,<,>=,<= are available. Wildcards % and _ are supported in the pattern string. % allows you to match any string of any length(including zero length) and _ allows you to match on a single character. The valid attributes are eventtype, monitor, monnode, application, component, id, serverity, message, rawdata, comments. Valid severity are: Informational, Warning, Critical.
- Operator descriptions:
== Select event where the attribute value is exactly this value. != Select event where the attribute value is not this specific value. =~ Select event where the attribute value matches this pattern string. Not work with severity. !~ Select event where the attribute value does not match this pattern string. Not work with severity. > Select event where the severity is higher than this value. Only work with severity. < Select event where the severity is lower than this value. Only work with severity. >= Select event where the severity is higher than this value(include). Only work with severity. <= Select event where the severity is lower than this value(include). Only work with severity.
Note: if the “val” or “operator” fields includes spaces or any other characters that will be parsed by shell, the “attr<operator>val” needs to be quoted. If the operator is ”!~”, the “attr<operator>val” needs to be quoted using single quote.
-o specifies montype, it can be p or e. p means performance, e means events.
- To show summary data about PctRealMemFree and PctTotalTimeIdle of cluster in last 6 minutes, enter:
monshow rmcmon -s -a PctRealMemFree,PctTotalTimeIdle -t 6
- To show all data of node1 and node2, enter:
monshow rmcmon node1,node2
- To show summary data of nodes which managed by servicenode1, enter:
monshow rmcmon servicenode1 -s
- To show RMC event with severity Critical, enter:
monshow rmcmon -w severity==Critical
/opt/xcat/bin/monshow
monls(1)|monls.1, monstart(1)|monstart.1, monstop(1)|monstop.1, monadd(1)|monadd.1, monrm(1)|monrm.1, moncfg(1)|moncfg.1, mondecfg(1)|mondecfg.1
monstart.1¶
monstart - Starts a plug-in module to monitor the xCAT cluster.
This command is used to start a 3rd party software, (for example start the daemons), to monitor the xCAT cluster. The operation is performed on the management node and the service nodes of the given nodes. The operation will also be performed on the nodes if the -r option is specified.
name is the name of the monitoring plug-in module. For example, if the the name is called xxx, then the actual file name that the xcatd looks for is /opt/xcat/lib/perl/xCAT_monitoring/xxx.pm. Use monls -a command to list all the monitoring plug-in modules that can be used.
noderange is the nodes to be monitored. If omitted, all nodes will be monitored.
-h | --help Display usage message.
-r | --remote Specifies that the operation will also be performed on the nodes. For example, the3rd party monitoring software daemons on the nodes will also be started.
- To start gangliamon plug-in module (which interacts with Ganglia monitoring software) to monitor the xCAT cluster, enter:
monstart gangliamon -r
- To start xcatmon plug-in module to feed the node liveness status to xCAT’s nodelist table, enter:
monstart rmcmon
/opt/xcat/bin/monstart
monls(1)|monls.1, monstop(1)|monstop.1, monadd(1)|monadd.1, monrm(1)|monrm.1, moncfg(1)|moncfg.1, mondecfg(1)|mondecfg.1
monstop.1¶
monstop - Stops a monitoring plug-in module to monitor the xCAT cluster.
This command is used to stop a 3rd party software, (for example stop the daemons), from monitoring the xCAT cluster. The operation is performed on the management node and the service nodes of the given nodes. The operation will also be performed on the nodes if the -r option is specified.
name is the name of the monitoring plug-in module in the monitoring table. Use monls command to list all the monitoring plug-in modules that can be used.
noderange is the nodes to be stopped for monitoring. If omitted, all nodes will be stopped.
-h | -help Display usage message.
-r | --remote Specifies that the operation will also be performed on the nodes. For example, the3rd party monitoring software daemons on the nodes will also be stopped.
1.To stop gangliamon plug-in module (which interacts with Ganglia monitoring software) to monitor the xCAT cluster, enter:
monstop gangliamon
Please note that gangliamon must have been registered in the xCAT monitoring table. For a list of registered plug-in modules, use command monls.
/opt/xcat/bin/monstop
monls(1)|monls.1, monstart(1)|monstart.1, monadd(1)|monadd.1, monrm(1)|monrm.1, moncfg(1)|moncfg.1, mondecfg(1)|mondecfg.1
mysqlsetup.1¶
mysqlsetup - Sets up the MySQL or MariaDB database for xCAT to use.
mysqlsetup {-h | --help}
mysqlsetup {-v | --version}
mysqlsetup {-i | --init} [-f | --hostfile] [-o|-**-odbc] [-L|--LL] [**-V | --verbose]
mysqlsetup {-u | --update} [-f | --hostfile] [-o|-**-odbc] [-L|--LL] [**-V | --verbose]
mysqlsetup - Sets up the MySQL or MariaDB database (linux only for MariaDB) for xCAT to use. The mysqlsetup script is run on the Management Node as root after the MySQL code or MariaDB code has been installed. Before running the init option, the MySQL server should be stopped, if it is running. The xCAT daemon, xcatd, must be running, do not stop it. No xCAT commands should be run during the init process, because we will be migrating the xCAT database to MySQL or MariaDB and restarting the xcatd daemon as well as the MySQL daemon. For full information on all the steps that will be done, read the “Configure MySQL and Migrate xCAT Data to MySQL” sections in Setting_Up_MySQL_as_the_xCAT_DB Two passwords must be supplied for the setup, a password for the xcatadmin id and a password for the root id in the MySQL database. These will be prompted for interactively, unless the environment variables XCATMYSQLADMIN_PW and XCATMYSQLROOT_PW are set to the passwords for the xcatadmin id and root id in the database,resp. Note below we refer to MySQL but it works the same for MariaDB.
-h|--help
Displays the usage message.
-v|--version
Displays the release version of the code.
-V|--verbose
Displays verbose messages.
-i|--init
The init option is used to setup a xCAT database on an installed MySQL or MariaDB server for xCAT to use. The mysqlsetup script will check for the installed MariaDB server rpm first and will use MariaDB if it is installed. This involves creating the xcatdb database, the xcatadmin id, allowing access to the xcatdb database by the Management Node. It customizes the my.cnf configuration file for xcat and starts the MySQL server. It also backs up the current xCAT database and restores it into the newly setup xcatdb MySQL database. It creates the /etc/xcat/cfgloc file to point the xcatd daemon to the MySQL database and restarts the xcatd daemon using the database. On AIX, it additionally setup the mysql id and group and corrects the permissions in the MySQL install directories. For AIX, you should be using the MySQL rpms available from the xCAT website. For Linux, you should use the MySQL or MariaDB rpms shipped with the OS. You can chose the -f and/or the -o option, to run after the init.
-u|--update
To run the update option, you must first have run the -i option and have xcat successfully running on the MySQL database. You can chose the -f and/or the -o option, to update.
-f|--hostfile
This option runs during update, it will take all the host from the input file (provide a full path) and give them database access to the xcatdb in MySQL for the xcatadmin id. Wildcards and ipaddresses may be used. xCAT must have been previously successfully setup to use MySQL. xcatadmin and MySQL root password are required.
-o|--odbc
This option sets up the ODBC /etc/../odbcinst.ini, /etc/../odbc.ini and the .odbc.ini file in roots home directory will be created and initialized to run off the xcatdb MySQL database. See “Add ODBC Support” in Setting_Up_MySQL_as_the_xCAT_DB
-L|--LL
Additional database configuration specifically for the LoadLeveler product. See “Add ODBC Support” in Setting_Up_MySQL_as_the_xCAT_DB
*
XCATMYSQLADMIN_PW - the password for the xcatadmin id that will be assigned in the MySQL database.
*
XCATMYSQLROOT_PW - the password for the root id that will be assigned to the MySQL root id, if the script creates it. The password to use to run MySQL command to the database as the MySQL root id. This password may be different than the unix root password on the Management Node.
*
To setup MySQL for xCAT to run on the MySQL xcatdb database :
mysqlsetup -i
*
Add hosts from /tmp/xcat/hostlist that can access the xcatdb database in MySQL:
mysqlsetup -u -f /tmp/xcat/hostlist
Where the file contains a host per line, for example:
node1 1.115.85.2 10.%.%.% nodex.cluster.net
*
To setup the ODBC for MySQL xcatdb database access :
mysqlsetup -o
*
To setup MySQL for xCAT and add hosts from /tmp/xcat/hostlist and setup the ODBC in Verbose mode:
mysqlsetup -i -f /tmp/xcat/hostlist -o -V
nimnodecust.1¶
nimnodecust - Use this xCAT command to customize AIX/NIM standalone machines.
nimnodecust [-h|--help ]
nimnodecust [-V] -s lpp_source_name [-p packages] [-b installp_bundles] noderange [attr=val [attr=val ...]]
This xCAT command can be used to customize AIX/NIM standalone machines.
The software packages that you wish to install on the nodes must be copied to the appropriate directory locations in the NIM lpp_source resource provided by the “-s” option. For example, if the location of your lpp_source resource is “/install/nim/lpp_source/61lpp/” then you would copy RPM packages to “/install/nim/lpp_source/61lpp/RPMS/ppc” and you would copy your installp packages to “/install/nim/lpp_source/61lpp/installp/ppc”. Typically you would want to copy the packages to the same lpp_source that was used to install the node. You can find the location for an lpp_source with the AIX lsnim command. (Ex. “lsnim -l <lpp_source_name>”)
The packages you wish to install on the nodes may be specified with either a comma-separated list of package names or by a comma-separated list of installp_bundle names. The installp_bundle names are what were used when creating the corresponding NIM installp_bundle definitions. The installp_bundle definitions may also be used when installing the nodes.
A bundle file contains a list of package names. The RPMs must have a prefix of “R:” and the installp packages must have a prefix of “I:”. For example, the contents of a simple bundle file might look like the following.
# RPM
R:expect-5.42.1-3.aix5.1.ppc.rpm
R:ping-2.4b2_to-1.aix5.3.ppc.rpm
#installp
I:openssh.base
I:openssh.license
To create a NIM installp_bundle definition you can use the “nim -o define” operation. For example, to create a definition called “mypackages” for a bundle file located at “/install/nim/mypkgs.bnd” you could issue the following command.
“nim -o define -t installp_bundle -a server=master -a location=/install/nim/mypkgs.bnd mypackages”.
See the AIX documantation for more information on using installp_bundle files.
The xCAT nimnodecust command will automatically handle the distribution of the packages to AIX service nodes when using an xCAT hierachical environment.
attr=val [attr=val ...]
Specifies one or more “attribute equals value” pairs, separated by spaces. Attr=val pairs must be specified last on the command line. These are used to specify additional values that can be passed to the underlying NIM commands, (“nim -o cust...”). See the NIM documentation for valid “nim” command line options.
-b installp_bundle_names
A comma separated list of NIM installp_bundle names.
-h |--help
Display usage message.
-p package_names
A comma-separated list of software packages to install. Packages may be RPM or installp.
noderange
A set of comma delimited node names and/or group names. See the “noderange” man page for details on additional supported formats.
-V |--verbose
Verbose mode.
- Install the installp package “openssh.base.server” on an xCAT node named “node01”. Assume that the package has been copied to the NIM lpp_source resource called “61lppsource”.
nimnodecust -s 61lppsource -p openssh.base.server node01
- Install the product software contained in the two bundles called “llbnd” and “pebnd” on all AIX nodes contained in the xCAT node group called “aixnodes”. Assume that all the software packages have been copied to the NIM lpp_source resource called “61lppsource”.
nimnodecust -s 61lppsource -b llbnd,pebnd aixnodes
/opt/xcat/bin/nimnodecust
This command is part of the xCAT software product.
nimnodeset.1¶
nimnodeset - Use this xCAT command to initialize AIX/NIM standalone machines.
nimnodeset [-h|--help ]
nimnodeset [-V|--verbose] [-f|--force] [-i osimage_name] [-l location] [-p|--primarySN] [-b|--backupSN] noderange [attr=val [attr=val ...]]
This xCAT command can be used to initialize AIX/NIM standalone machines. Once this step is completed the either the xCAT rnetboot command or the rbootseq/rpower commands to initiate a network boot of the nodes.
If you are using xCAT service nodes the nimnodeset command will automatically determine the correct server(s) for the node and do the initialization on that server(s).
The osimage_name is the name of an xCAT osimage definition that contains the list of NIM resources to use when initializing the nodes. If the osimage_name is not provided on the command line the code checks the node definition for the value of the “provmethod” attribute (which is the name of an osimage definition). If the osimage_image is provided on the command line then the code will also set the “provmethod” attribute of the node definiions.
This command will also create a NIM resolv_conf resource to be used when installing the node. If a resolv_conf resource is not already included in the xCAT osimage definition and if the “domain” and “nameservers” values are set then a new NIM resolv_conf resource will be created and allocated to the nodes.
The “domain” and “nameservers” attributes can be set in either the xCAT “network” definition used by the nodes or in the xCAT cluster “site” definition. The setting in the “network” definition will take priority.
The “search” field of the resolv.conf file will contain a list all the domains listed in the xCAT network definitions and the xCAT site definiton.
The “nameservers” value can either be set to a specific IP address or the “<xcatmaster>” key word. The “<xcatmaster>” key word means that the value of the “xcatmaster” attribute of the node definition will be used in the /etc/resolv.conf file. (I.e. The name of the install server as known by the node.)
You can set the “domain” and “nameservers” attributes by using the chdef command. For example:
chdef -t network -o clstr_net domain=cluster.com nameservers=<xcatmaster>
If the “domain” and “nameservers” attributes are not set in either the nodes “network” definition or the “site” definition then no new NIM resolv_conf resource will be created.
You can specify additional attributes and values using the “attr=val” command line option. This information will be passed on to the underlying call to the NIM “nim -o bos_inst” command. See the NIM documentation for information on valid command line options for the nim command. The “attr” must correspond to a NIM attribute supported for the NIM “bos_inst” operation. Information provided by the “attr=val” option will take precedence over the information provided in the osimage definition.
The force option can be used to reinitialize a node if it already has resources allocated or it is in the wrong NIM state. This option will reset the NIM node and deallocate resources before reinititializing.
This command will also create a NIM script resource to enable the xCAT support for user-provided customization scripts.
After the nimnodeset command completes you can use the lsnim command to check the NIM node definition to see if it is ready for booting the node. (“lsnim -l <nim_node_name>”).
You can supply your own scripts to be run on the management node or on the service node (if their is hierarchy) for a node during the nimnodeset command. Such scripts are called prescripts. They should be copied to /install/prescripts dirctory. A table called prescripts is used to specify the scripts and their associated actions. The scripts to be run at the beginning of the nimnodeset command are stored in the ‘begin’ column of prescripts table. The scripts to be run at the end of the nimnodeset command are stored in the ‘end’ column of prescripts table. Please run ‘tabdump prescripts -d’ command for details. An example for the ‘begin’ or the ‘end’ column is: standalone:myscript1,myscript2. The following two environment variables will be passed to each script: NODES contains all the names of the nodes that need to run the script for and ACTION contains the current nodeset action, in this case “standalone”. If #xCAT setting:MAX_INSTANCE=number is specified in the script, the script will get invoked for each node in parallel, but no more than number of instances will be invoked at at a time. If it is not specified, the script will be invoked once for all the nodes.
attr=val [attr=val ...]
Specifies one or more “attribute equals value” pairs, separated by spaces. Attr= val pairs must be specified last on the command line. These are used to specify additional values that can be passed to the underlying NIM commands, (“nim -o bos_inst ...”). See the NIM documentation for valid “nim” command line options. Note that you may specify multiple “script” and “installp_bundle” values by using a comma seperated list. (ex. “script=ascript,bscript”).
-b|--backupSN
When using backup service nodes only update the backup. The default is to update both the primary and backup service nodes
-f |--force
Use the force option to reinitialize the NIM machines.
-h |--help
Display usage message.
-i image_name
The name of an existing xCAT osimage definition.
-l|--location
The directory location to use when creating new NIM resolv_conf resources. The d efault location is /install/nim.
-p|--primarySN
When using backup service nodes only update the primary. The default is to update both the primary and backup service nodes.
noderange
A set of comma delimited node names and/or group names. See the “noderange” man page for details on additional supported formats.
-V |--verbose
Verbose mode.
- Initialize an xCAT node named “node01”. Use the xCAT osimage named “61gold” to install the node.
nimnodeset -i 61gold node01
- Initialize all AIX nodes contained in the xCAT node group called “aixnodes” using the image definitions pointed to by the “provmethod” attribute of the xCAT node definitions.
nimnodeset aixnodes
- Initialize an xCAT node called “node02”. Include installp_bundle resources that are not included in the osimage definition. This assumes the NIM installp_bundle resources have already been created.
nimnodeset -i 611image node02 installp_bundle=sshbundle,addswbundle
/opt/xcat/bin/nimnodeset
This command is part of the xCAT software product.
mknimimage(1)|mknimimage.1, rnetboot(1)|rnetboot.1
nodeaddunmged.1¶
nodeaddunmged - Create a unmanaged node.
nodeaddunmged [-h| -**-help | -v | -**-version]
nodeaddunmged hostname=<node-name> ip=<ip-address>
The nodeaddunmged command adds an unmanaged node to the __Unmanaged group. You can specify the node name and IP address of the node.
-h|--help
Display usage message.
-v|--version
Command Version.
hostname=<node-name>
Sets the name of the new unmanaged node, where <node-name> is the name of the node.
ip=<ip-address>
Sets the IP address of the unmanaged node, where <ip-address> is the IP address of the new node in the form xxx.xxx.xxx.xxx
To add an unmanaged node, use the following command: nodeaddunmged hostname=unmanaged01 ip=192.168.1.100
nodech.1¶
nodech - Changes nodes’ attributes in the xCAT cluster database.
nodech noderange table.column=value [...]
nodech {-d | --delete} noderange table [...]
nodech {-v | --version}
nodech [-? | -h | --help]
The nodech command changes the specified attributes for the given nodes. Normally, the given value will completely replace the current attribute value. But if ”,=” is used instead of “=”, the specified value will be prepended to the attribute’s comma separated list, if it is not already there. If “^=” is used, the specified value will be removed from the attribute’s comma separated list, if it is there. You can also use “^=” and ”,=” in the same command to essentially replace one item in the list with another. (See the Examples section.)
Additionally, as in nodels, boolean expressions can be used to further limit the scope of nodech from the given noderange. The operators supported are the same as nodels (=~, !~, ==, and !=).
With these operators in mind, the unambiguous assignment operator is ‘=@’. If you need, for example, to set the nodelist.comments to =foo, you would have to do nodech n1 nodelist.comments=@=foo.
See the xcatdb man page for an overview of each table.
The nodech command also supports some short cut names as aliases to common attributes. See the nodels man page for details.
-d|--delete
Delete the nodes’ row in the specified tables.
-v|--version
Command Version.
-?|-h|--help
Display usage message.
*
To update nodes in noderange node1-node4 to be in only group all:
nodech node1-node4 groups=all
*
To put all nodes with nodepos.rack value of 2 into a group called rack2:
nodech all nodepos.rack==2 groups,=rack2
*
To add nodes in noderange node1-node4 to the nodetype table with os=rhel5:
nodech node1-node4 groups=all,rhel5 nodetype.os=rhel5
*
To add node1-node4 to group1 in addition to the groups they are already in:
nodech node1-node4 groups,=group1
*
To put node1-node4 in group2, instead of group1:
nodech node1-node4 groups^=group1 groups,=group2
/opt/xcat/bin/nodech
nodels(1)|nodels.1, nodeadd(8)|nodeadd.8, noderange(3)|noderange.3
nodechmac.1¶
nodechmac - Updates the MAC address for a node.
The nodechmac command changes the MAC address for provisioned node’s network interface.
You can use this command to keep an existing node configuration. For example, if an existing node has hardware problems, the replacement node can use the old configurations. By using the nodechmac command, the node name and network settings of the old node can be used by the new node.
-h|--help
Display usage message.
-v|--version
Command Version.
node-name
Specifies the name of the node you want to update, where <node-name> is the node that is updated.
mac=<mac-address
Sets the new MAC address for the NIC used by the provisioning node, where <mac-address> is the NICs new MAC address.
You can update the MAC address for a node, by using the following command: nodechmac compute-000 mac=2F:3C:88:98:7E:01
nodechprofile.1¶
nodechprofile - updates a profile used by a node
nodechprofile [-h| -**-help | -v | -**-version]
nodechprofile <noderange> [imageprofile=<image-profile>] [networkprofile=<network-profile>] [hardwareprofile=<hardware-profile>]
The nodechprofile command updates the profiles used by a node, including: the image profile, network profile, and hardware management profile.
If you update the image profile for a node, the operating system and provisioning settings for the node are updated.
If you update the network profile, the IP address and network settings for the node are updated.
If you update the hardware management profile, the hardware settings for the node are updated.
After nodes’ hardware profile or image profile are updated, the status for each node is changed to “defined”. A node with a “defined” status must be reinstalled
After nodes’ network profile updated, the status for nodes is not changed. You’ll need to run noderegenips to re-generate the nodes’ IP address and nodes’ status may also be updated at this stage.
-h|--help
Display usage message.
-v|--version
Command Version.
noderange
The nodes to be removed.
imageprofile=<image-profile>
Sets the new image profile name used by the node, where <image-profile> is the new image profile. An image profile defines the provisioning method, OS information, kit information, and provisioning parameters for a node. If the “__ImageProfile_imgprofile” group already exists in the nodehm table, then “imgprofile” is used as the image profile name.
networkprofile=<network-profile>
Sets the new network profile name used by the node, where <network-profile> is the new network profile. A network profile defines the network, NIC, and routes for a node. If the “__NetworkProfile_netprofile” group already exists in the nodehm table, then “netprofile” is used as the network profile name.
hardwareprofile=<hardware-profile>
Sets the new hardware profile name used by the node, where <hardware-profile> is the new hardware management profile used by the node. If a “__HardwareProfile_hwprofile” group exists, then “hwprofile” is the hardware profile name. A hardware profile defines hardware management related information for imported nodes, including: IPMI, HMC, CEC, CMM.
To change the image profile to rhels6.3_packaged for compute nodes compute-000 and compute-001, use the following command:
nodechprofile compute-000,compute-001 imageprofile=rhels6.3_packaged
To change all of the profiles for compute node compute-000, enter the following command:
nodechprofile compute-000 imageprofile=rhels6.3_packaged networkprofile=default_cn hardwareprofile=default_ipmi
nodepurge(1)|nodepurge.1, noderefresh(1)|noderefresh.1, nodeimport(1)|nodeimport.1, noderange(3)|noderange.3
nodediscoverdef.1¶
nodediscoverdef - Define the undefined discovery request to a predefined xCAT node, or clean up the discovery entries from the discoverydata table (which can be displayed by nodediscoverls command)
nodediscoverdef -u uuid -n node
nodediscoverdef -r -u uuid
nodediscoverdef -r -t {seq | profile | switch | blade | manual | undef | all}
nodediscoverdef [-h | --help | -v | --version]
The nodediscoverdef command defines the discovery entry from the discoverydata table to a predefined xCAT node. The discovery entry can be displayed by nodediscoverls command.
The options -u and -n have to be used together to define a discovery request to a node.
The nodediscoverdef command also can be used to clean up the discovery entries from the discoverydata table.
The option -r is used to remove discovery entries. If working with -u, the specific entry which uuid specified by -u will be removed.
You also can use the -r -t option to limit that only remove the nodes that were discovered in a particular method of discovery.
-t seq|profile|switch|blade|manual|undef|all
Specify the nodes that have been discovered by the specified discovery method:
*
seq - Sequential discovery (started via nodediscoverstart noderange=<noderange> ...).*
profile - Profile discovery (started via nodediscoverstart networkprofile=<network-profile> ...).*
switch - Switch-based discovery (used when the switch and switches tables are filled in).*
blade - Blade discovery (used for IBM Flex blades).*
manual - Manually discovery (used when defining node by nodediscoverdef command).*
undef - Display the nodes that were in the discovery pool, but for which xCAT has not yet received a discovery request.*
all - All discovered nodes.
-n node
The xCAT node that the discovery entry will be defined to.
-r
Remove the discovery entries from discoverydata table.
-u uuid
The uuid of the discovered entry.
-h|--help
Display usage message.
-v|--version
Command version.
1
Define the discovery entry which uuid is 51E5F2D7-0D59-11E2-A7BC-3440B5BEDBB4 to node node1
nodediscoverdef -u 51E5F2D7-0D59-11E2-A7BC-3440B5BEDBB4 -n node1
Defined [51E5F2D7-0D59-11E2-A7BC-3440B5BEDBB4] to node node1.
2
Remove the discovery entry which uuid is 51E5F2D7-0D59-11E2-A7BC-3440B5BEDBB4 from the discoverydata table
nodediscoverdef -r -u 51E5F2D7-0D59-11E2-A7BC-3440B5BEDBB4
Removing discovery entries finished.
3
Remove the discovery entries which discover type is seq from the discoverydata table
nodediscoverdef -r -t seq
Removing discovery entries finished.
nodediscoverstart(1)|nodediscoverstart.1, nodediscoverstatus(1)|nodediscoverstatus.1, nodediscoverstop(1)|nodediscoverstop.1, nodediscoverls(1)|nodediscoverls.1
nodediscoverls.1¶
nodediscoverls - List the discovered nodes
nodediscoverls [-t seq | profile | switch | blade | manual | undef | all] [-l]
nodediscoverls [-u uuid] [-l]
nodediscoverls [-h | --help | -v | --version]
The nodediscoverls command lists nodes that have recently been discovered. If discovery is currently in progress (i.e. nodediscoverstart has been run, but nodediscoverstop has not been), then nodediscoverls will list the nodes that have been discovered so far in this session. If discovery is not currently in progress, nodediscoverls will list all of the nodes that were discovered in the last discovery session.
You can use the -t option to limit the output to just the nodes that were discovered in a particular method of discovery.
-t seq|profile|switch|blade|manual|undef|all
Display the nodes that have been discovered by the specified discovery method:
*
seq - Sequential discovery (started via nodediscoverstart noderange=<noderange> ...).*
profile - Profile discovery (started via nodediscoverstart networkprofile=<network-profile> ...).*
switch - Switch-based discovery (used when the switch and switches tables are filled in).*
blade - Blade discovery (used for IBM Flex blades).*
manual - Manually discovery (used when defining node by nodediscoverdef command).*
undef - Display the nodes that were in the discovery pool, but for which xCAT has not yet received a discovery request.*
all - All discovered nodes.
-l
Display more detailed information about the discovered nodes.
-u uuid
Display the discovered node that has this uuid.
-h|--help
Display usage message.
-v|--version
Command version.
1
Display the discovered nodes when sequential discovery is running:
nodediscoverls
UUID NODE METHOD MTM SERIAL 51E5F2D7-0D59-11E2-A7BC-3440B5BEDBB2 distest1 sequential 786310X 1052EF2 51E5F2D7-0D59-11E2-A7BC-3440B5BEDBB3 distest2 sequential 786310X 1052EF3
2
Display the nodes that were in the discovery pool, but for which xCAT has not yet received a discovery request:
nodediscoverls -t undef
UUID NODE METHOD MTM SERIAL 51E5F2D7-0D59-11E2-A7BC-3440B5BEDBB0 undef undef 786310X 1052EF0 51E5F2D7-0D59-11E2-A7BC-3440B5BEDBB1 undef undef 786310X 1052EF1
3
Display all the discovered nodes:
nodediscoverls -t all
UUID NODE METHOD MTM SERIAL 51E5F2D7-0D59-11E2-A7BC-3440B5BEDBB0 undef undef 786310X 1052EF0 51E5F2D7-0D59-11E2-A7BC-3440B5BEDBB1 undef undef 786310X 1052EF1 51E5F2D7-0D59-11E2-A7BC-3440B5BEDBB2 distest1 sequential 786310X 1052EF2 51E5F2D7-0D59-11E2-A7BC-3440B5BEDBB3 distest2 sequential 786310X 1052EF3
4
Display the discovered node whose uuid is 51E5F2D7-0D59-11E2-A7BC-3440B5BEDBB2, with detailed information:
nodediscoverls -u 51E5F2D7-0D59-11E2-A7BC-3440B5BEDBB2 -l
Object uuid: 51E5F2D7-0D59-11E2-A7BC-3440B5BEDBB2 node=distest1 method=sequential discoverytime=03-31-2013 17:05:12 arch=x86_64 cpucount=32 cputype=Intel(R) Xeon(R) CPU E5-2690 0 @ 2.90GHz memory=198460852 mtm=786310X serial=1052EF2 nicdriver=eth0!be2net,eth1!be2net nicipv4=eth0!10.0.0.212/8 nichwaddr=eth0!34:40:B5:BE:DB:B0,eth1!34:40:B5:BE:DB:B4 nicpci=eth0!0000:0c:00.0,eth1!0000:0c:00.1 nicloc=eth0!Onboard Ethernet 1,eth1!Onboard Ethernet 2 niconboard=eth0!1,eth1!2 nicfirm=eth0!ServerEngines BE3 Controller,eth1!ServerEngines BE3 Controller switchname=eth0!c909f06sw01 switchaddr=eth0!192.168.70.120 switchdesc=eth0!IBM Flex System Fabric EN4093 10Gb Scalable Switch, flash image: version 7.2.6, boot image: version 7.2.6 switchport=eth0!INTA2
nodediscoverstart(1)|nodediscoverstart.1, nodediscoverstatus(1)|nodediscoverstatus.1, nodediscoverstop(1)|nodediscoverstop.1, nodediscoverdef(1)|nodediscoverdef.1
nodediscoverstart.1¶
nodediscoverstart - starts the node discovery process
nodediscoverstart [-h | --help | -v | --version]
Sequential Discovery Specific:
nodediscoverstart noderange=noderange [hostiprange=imageprofile] [bmciprange=bmciprange] [groups=groups] [rack=rack] [chassis=chassis] [height=height] [unit=unit] [osimage=<osimagename>] [-n|-**-dns] [-s|--skipbmcsetup] [**-V|--verbose]
Profile Discovery Specific:
nodediscoverstart networkprofile=network-profile imageprofile=image-profile hostnameformat=nost-name-format [hardwareprofile=hardware-profile] [groups=node-groups] [rack=rack-name] [chassis=chassis-name] [height=rack-server-height] [unit=rack-server-unit-location] [rank=rank-num]
The nodediscoverstart command starts either the Sequential Discovery or Profile Discovery process. They can not both be running at the same time.
Sequential Discovery Specific:
This is the simplest discovery approach. You only need to specify the noderange, hostiprange and bmciprange that should be given to nodes that are discovered. (If you pre-define the nodes (via nodeadd or mkdef) and specify their host and BMC IP addresses, then you only need to specify the noderange to the nodediscoverstart command.) Once you have run nodediscoverstart, then physically power on the nodes in the sequence that you want them to receive the node names and IPs, waiting a short time (e.g. 30 seconds) between each node.
Profile Discovery Specific:
This is the PCM discovery approach. networkprofile, imageprofile, hostnameformat arguments must be specified to start the Profile Discovery. All nodes discovered by this process will be associated with specified profiles and rack/chassis/unit locations.
When the nodes are discovered, PCM updates the affected configuration files on the management node automatically. Configuration files include the /etc/hosts service file, DNS configuration, and DHCP configuration. Kit plug-ins are automatically triggered to update kit related configurations and services.
When you power on the nodes, they PXE boot and DHCP/TFTP/HTTP on the management node give each node the xCAT genesis boot image, which inventories the node hardware and sends data to the management node. There, either the sequential discovery process or the profile discovery process assigns node attributes and defines the node in the the database.
noderange=noderange
The set of node names that should be given to nodes that are discovered via the Sequential Discovery method. This argument is required to Sequential Discovery. Any valid xCAT noderange is allowed, e.g. node[01-10].
hostiprange=ip range
The ip range which will be assigned to the host of new discovered nodes in the Sequential Discovery method. The format can be: start_ip-end_ip or noderange, e.g. 192.168.0.1-192.168.0.10 or 192.168.0.[1-10].
bmciprange=ip range
The ip range which will be assigned to the bmc of new discovered nodes in the Sequential Discovery method. The format can be: start_ip-end_ip or noderange, e.g. 192.168.1.1-192.168.1.10 or 192.168.1.[1-10].
imageprofile=image-profile
Sets the new image profile name used by the discovered nodes in the Profile Discovery method. An image profile defines the provisioning method, OS information, kit information, and provisioning parameters for a node. If the “__ImageProfile_imgprofile” group already exists in the nodehm table, then “imgprofile” is used as the image profile name.
networkprofile=network-profile
Sets the new network profile name used by the discovered nodes in the Profile Discovery method. A network profile defines the network, NIC, and routes for a node. If the “__NetworkProfile_netprofile” group already exists in the nodehm table, then “netprofile” is used as the network profile name.
hardwareprofile=hardware-profile
Sets the new hardware profile name used by the discovered nodes in the Profile Discovery method. If a “__HardwareProfile_hwprofile” group exists, then “hwprofile” is the hardware profile name. A hardware profile defines hardware management related information for imported nodes, including: IPMI, HMC, CEC, CMM.
hostnameformat=nost-name-format
Sets the node name format for all discovered nodes in the Profile Discovery method. The two types of formats supported are prefix#NNNappendix and prefix#RRand#NNappendix, where wildcard #NNN and #NN are replaced by a system generated number that is based on the provisioning order. Wildcard #RR represents the rack number and stays constant.
For example, if the node name format is compute-#NN, the node name is generated as: compute-00, compute-01, ..., compute-99. If the node name format is blade#NNN-x64, the node name is generated as: blade001-x64, blade002-x64, ..., blade999-x64
For example, if the node name format is compute-#RR-#NN and the rack number is 2, the node name is generated as: compute-02-00, compute-02-01, ..., compute-02-99. If node name format is node-#NN-in-#RR and rack number is 1, the node name is generated as: node-00-in-01, node-01-in-01, ..., node-99-in-01
groups=node-groups
Sets the node groups that the discovered nodes should be put in for either the Sequential Discovery or Profile Discovery methods, where node-group is a comma-separated list of node groups.
rack=rack-name>
Sets the rack name where the node is located for either the Sequential Discovery or Profile Discovery methods.
chasiss=chassis-name
Sets the chassis name that the Blade server or PureFlex blade is located in, for either the Sequential Discovery or Profile Discovery methods. This option is used for the Blade server and PureFlex system only. You cannot specify this option with the rack option.
height=rack-server-height
Sets the height of a rack-mounted server in U units for either the Sequential Discovery or Profile Discovery methods. If the rack option is not specified, the default value is 1.
unit=rack-server-unit-location
Sets the start unit value for the node in the rack, for either the Sequential Discovery or Profile Discovery methods. This option is for a rack server only. If the unit option is not specified, the default value is 1
rank=rank-num
Specifies the starting rank number that is used in the node name format, for the Profile Discovery method. The rank number must be a valid integer between 0 and 254. This option must be specified with nodenameformat option. For example, if your node name format is compute-#RR-#NN. The rack’s number is 2 and rank is specified as 5, the node name is generated as follows: compute-02-05, compute-02-06, ..., compute-02-99.
osimage=osimagename
Specifies the osimage name that will be associated with the new discovered node, the os provisioning will be started automatically at the end of the discovery process.
-n|--dns
Specifies to run makedns <nodename> for any new discovered node. This is useful mainly for non-predefined configuration, before running the “nodediscoverstart -n”, the user needs to run makedns -n to initialize the named setup on the management node.
-s|--skipbmcsetup
Specifies to skip the bmcsetup during the sequential discovery process, if the bmciprange is specified with nodediscoverstart command, the BMC will be setup automatically during the discovery process, if the user does not want to run bmcsetup, could specify the “-s|–skipbmcsetup” with nodediscoverstart command to skip the bmcsetup.
-V|--verbose
Enumerates the free node names and host/bmc ips that are being specified in the ranges given. Use this option with Sequential Discovery to ensure that you are specifying the ranges you intend.
-h|--help
Display usage message.
-v|--version
Command Version.
1
Sequential Discovery: To discover nodes with noderange and host/bmc ip range:
nodediscoverstart noderange=n[1-10] hostiprange=‘172.20.101.1-172.20.101.10’ bmciprange=‘172.20.102.1-172.20.102.10’ -V
Sequential Discovery: Started: Number of free node names: 10 Number of free host ips: 10 Number of free bmc ips: 10 ------------------------------------Free Nodes------------------------------------ NODE HOST IP BMC IP n01 172.20.101.1 172.20.102.1 n02 172.20.101.2 172.20.102.2 ... ... ...
2
Profile Discovery: To discover nodes using the default_cn network profile and the rhels6.3_packaged image profile, use the following command:
nodediscoverstart networkprofile=default_cn imageprofile=rhels6.3_packaged hostnameformat=compute#NNN
nodediscoverstop(1)|nodediscoverstop.1, nodediscoverls(1)|nodediscoverls.1, nodediscoverstatus(1)|nodediscoverstatus.1
nodediscoverstatus.1¶
nodediscoverstatus - gets the node discovery process status
nodediscoverstatus [-h | --help | -v | --version]
The nodediscoverstatus command detects if the sequential or profile node discovery process is currently running, i.e. nodediscoverstarthas been run, but nodediscoverstop has not.
To determine if there are some nodes discovered and the discovered nodes’ status, enter the following command:
nodediscoverstatus
nodediscoverstart(1)|nodediscoverstart.1, nodediscoverls(1)|nodediscoverls.1, nodediscoverstatus(1)|nodediscoverstop.1
nodediscoverstop.1¶
nodediscoverstop - stops the node discovery process.
nodediscoverstop [-h | --help | -v | --version]
The nodediscoverstop command stops the sequential or profile node discovery process. Once this command has been run, newly discovered nodes will not be assigned node names and attributes automatically via the sequential or profile discovery process.
nodediscoverstop
nodediscoverstart(1)|nodediscoverstart.1, nodediscoverls(1)|nodediscoverls.1, nodediscoverstatus(1)|nodediscoverstatus.1
nodegrpch.1¶
nodegrpch - Changes attributes at the group level in the xCAT cluster database.
nodegrpch group1,group2,... table.column=value [...]
nodegrpch {-v | --version}
nodegrpch [-? | -h | --help]
The nodegrpch command is similar to the nodech command, but ensures that the parameters are declared at the group level rather than the node specific level, and clears conflicting node specific overrides of the specified groups. Using table.column=value will do a verbatim assignment. If ”,=” is used instead of “=”, the specified value will be prepended to the attribute’s comma separated list, if it is not already there. If “^=” is used, the specified value will be removed from the attribute’s comma separated list, if it is there. You can also use “^=” and ”,=” in the same command to essentially replace one item in the list with another. (See the Examples section.)
With these operators in mind, the unambiguous assignment operator is ‘=@’. If you need, for example, to set the nodehm.comments to =foo, you would have to do nodegrpch group1 nodehm.comments=@=foo.
See the xcatdb man page for an overview of each table.
The nodegrpch command also supports some short cut names as aliases to common attributes. See the nodels man page for details.
*
To declare all members of ipmi group to have nodehm.mgt be ipmi
** nodegrpch** ipmi nodehm.mgt=ipmi
/opt/xcat/bin/nodegrpch
nodech(1)|nodech.1, nodels(1)|nodels.1, nodeadd(8)|nodeadd.8, noderange(3)|noderange.3
nodeimport.1¶
nodeimport - Create profiled nodes by importing hostinfo file.
nodeimport [-h| -**-help | -v | -**-version]
nodeimport file=<hostinfo-filename> networkprofile=<network-profile> imageprofile=<image-profile> hostnameformat=<node-name-format> [hardwareprofile=<hardware-profile>] [groups=<node-groups>]
The nodeimport command creates nodes by importing a hostinfo file which is following stanza format. In this hostinfo file, we can define node’s hostname, ip, mac, switch name, switch port and host location infomation like rack, chassis, start unit, server height...etc
After nodes imported, the configuration files related with these nodes will be updated automatically. For example: /etc/hosts, dns configuration, dhcp configuration. And the kits node plugins will also be triggered automatically to update kit related configuration/services.
-h|--help
Display usage message.
-v|--version
Command Version.
file=<nodeinfo-filename>
Specifies the node information file, where <nodeinfo-filename> is the full path and file name of the node information file.
imageprofile=<image-profile>
Sets the new image profile name used by the node, where <image-profile> is the new image profile. An image profile defines the provisioning method, OS information, kit information, and provisioning parameters for a node. If the “__ImageProfile_imgprofile” group already exists in the nodehm table, then “imgprofile” is used as the image profile name.
networkprofile=<network-profile>
Sets the new network profile name used by the node, where <network-profile> is the new network profile. A network profile defines the network, NIC, and routes for a node. If the “__NetworkProfile_netprofile” group already exists in the nodehm table, then “netprofile” is used as the network profile name.
hardwareprofile=<hardware-profile>
Sets the new hardware profile name used by the node, where <hardware-profile> is the new hardware management profile used by the node. If a “__HardwareProfile_hwprofile” group exists, then “hwprofile” is the hardware profile name. A hardware profile defines hardware management related information for imported nodes, including: IPMI, HMC, CEC, CMM.
hostnameformat=<host-name-format>
Sets the node name format for all nodes discovered, where <node-name-format> is a supported format. The two types of formats supported are prefix#NNNappendix and prefix#RRand#NNappendix, where wildcard #NNN and #NN are replaced by a system generated number that is based on the provisioning order. Wildcard #RR represents the rack number and stays constant.
For example, if the node name format is compute-#NN, the node name is generated as: compute-00, compute-01, ... , compute-99. If the node name format is blade#NNN-x64, the node name is generated as: blade001-x64, blade002-x64, ... , blade999-x64
For example, if the node name format is compute-#RR-#NN and the rack number is 2, the node name is generated as: compute-02-00, compute-02-01, ..., compute-02-99. If node name format is node-#NN-in-#RR and rack number is 1, the node name is generated as: node-00-in-01, node-01-in-01, ... , node-99-in-01
groups=<node-groups>
Sets the node groups that the imported node belongs to, where <node-group> is a comma-separated list of node groups.
0 The command completed successfully.
1 An error has occured while validating parameters.
2 An error has occured while parsing hostinfo file.
To import nodes using a profile, follow the following steps:
- Find all node groups and profiles, run the following command “tabdump nodegroups”. For detailed profile information run “lsdef -t group <groupname>”. Example of detailed profile information:
# tabdump nodegroup
#groupname,grouptype,members,membergroups,wherevals,comments,disable
"compute","static",,,,,
"__HardwareProfile_default_ipmi","static","static",,,,
"__NetworkProfile_default_mn","static","static",,,,
"__NetworkProfile_default_cn","static",,,,,
"__ImageProfile_rhels6.2-x86_64-install-compute","static","static",,,,
# lsdef -t group __NetworkProfile_default_cn
Object name: __NetworkProfile_default_cn
grouptype=static
installnic=eth0
members=compute-000,compute-001
netboot=xnba
nichostnamesuffixes=eth0:-eth0
nicnetworks=eth0:provision
nictypes=eth0:Ethernet
primarynic=eth0
- Prepare a node information file.
Example of a node information file, a blade and a rack server defined:
# hostinfo begin
# This entry defines a blade.
__hostname__:
mac=b8:ac:6f:37:59:24
ip=192.168.1.20
chassis=chassis01
# This entry defines a rack server.
__hostname__:
mac=b8:ac:6f:37:59:25
ip=192.168.1.20
rack=rack01
height=1
unit=2
# hostinfo end.
Another example of a node infomation file, a PureFlex X/P node defined:
# hostinfo begin
# To define a PureFlex P/X node, chassis and slot id must be specified.
# The chassis must be a PureFlex chassis.
__hostname__:
mac=b8:ac:6f:37:59:25
chassis=cmm01
slotid=1
# hostinfo end.
Example of a node information file, a switch auto discovery node defined:
# hostinfo begin
# This entry defines a blade.
__hostname__:
switches=eth0!switch1!1,eth0!switch2!1!eth1
Example of a node information file that specifies a CEC-based rack-mounted Power node that uses direct FSP management:
# Node information file begins
# This entry defines a Power rack-mount node.
__hostname__:
mac=b8:ac:6f:37:59:28
cec=mycec
__hostname__:
mac=b8:ac:6f:37:59:28
cec=mycec
lparid=2
# Node information file ends.
Example of a node information file that specifies a PowerKVM Guest node that uses KVM management:
# Node information file begins
# This entry defines a PowerKVM Guest node.
# Make sure the node 'vm01' is already created on Hypervisor
vm01:
mac=b8:ef:3f:28:31:15
vmhost=pkvm1
# Node information file ends.
The node information file includes the following items:
__hostname__: This is a mandatory item.
Description: The name of the node, where __hostname__ is automatically generated by the node name format. You can also input a fixed node name, for example “compute-node”.
mac=<mac-address> This is a mandatory item.
Description: Specify the MAC address for the NIC used by the provisionging node, where <mac-address> is the NICs MAC address.
switches=<nic-name!switch-name!switch-port> This is a mandatory item, when define switch, switchport and node nic name relationship.
Description: Specify nic name, switch name and switch port to define node and switch relationship. We can define multi nic-switch-port relations here, looks like: switches=eth0!switch1!1,eth1!switch1,2
slotid=<slot-id> This is a mandatory item while define a PureFlex node.
Description: The node position in the PureFlex Chassis.
cec=<cec-name> This is a mandatory option for defining Power rack-mounted nodes.
Description: Specifies the name of a Power rack-mount central electronic complex (CEC).
lparid=<lpar-id> This is a optional option for defining Power rack-mounted nodes.
Description: Specifies the LPAR ID of a Power rack-mounted node, where <lpar-id> is the ID number. The default value is 1 if it is not defined.
ip=<ip-address> This is an optional item.
Description: Specify the IP address used for provisioning a node, where <ip-address> is in the form xxx.xxx.xxx.xxx. If this item is not included, the IP address used to provision the node is generated automatically according to the Network Profile used by the node.
nicips=<nics-ip> This is an optional item.
Description: Lists the IP address for each network interface configuration (NIC) used by the node, excluding the provisioning network, where <nics-ip> is in the form <nic1>!<nic-ip1>,<nic2>!<nic-ip2>,.... For example, if you have 2 network interfaces configured, the nicips attribute should list both network interfaces: nicips=eth1!10.10.10.11,bmc!192.168.10.3. If the nicips attribute is not specified, the IP addresses are generated automatically according to the network profile.
rack=<rack-name> This is an optional item.
Description: node location info. Specify the rack name which this node will be placed into. If not specify this item, there will be no node location info set for this node. this item must be specified together with height + unit.
chasiss=<chassis-name> This is an optional item.
Description: node location info, for blade(or PureFlex) only. Specify the chasiss name which this blade will be placed into. this item can not be specified together with rack.
height=<chassis-height> This is an optional item.
Description: node location info, for rack server only. Specify the server height number, in U. this item must be specified together with rack and unit.
unit=<rack-server-unit-location> This is an optional item.
Description: node location info, for rack server only. Specify the node’s start unit number in rack, in U. this item must be specified together with rack and height.
vmhost=<PowerKVM Hypervisior Host Name> This is a mandatory option for defining PowerKVM Guest nodes.
Description: Specifies the vmhost of a Power KVM Guest node, where <vmhost> is the host name of PowerKVM Hypervisior.
- Import the nodes, by using the following commands. Note: If we want to import PureFlex X/P nodes, hardware profile must be set to a PureFlex hardware type.
nodeimport file=/root/hostinfo.txt networkprofile=default_cn imageprofile=rhels6.3_packaged hostnameformat=compute-#NNN
- After importing the nodes, the nodes are created and all configuration files used by the nodes are updated, including: /etc/hosts, DNS, DHCP.
- Reboot the nodes. After the nodes are booted they are provisioned automatically.
nodepurge(1)|nodepurge.1, nodechprofile(1)|nodechprofile.1, noderefresh(1)|noderefresh.1
nodels.1¶
nodels - lists the nodes, and their attributes, from the xCAT database.
nodels [noderange] [-b | --blame] [-H | --with-fieldname] [-S] [table.column | shortname] [...]
nodels [noderange] [-H | --with-fieldname] [table]
nodels [-? | -h | --help | -v | --version]
The nodels command lists the nodes specified in the node range. If no noderange is provided, then all nodes are listed.
Additional attributes of the nodes will also be displayed if the table names and attribute names are specified after the noderange in the form: table.column . A few shortcut names can also be used as aliases to common attributes:
groups
nodelist.groups
tags
nodelist.groups
mgt
nodehm.mgt
nodels can also select based on table value criteria. The following operators are available:
==
Select nodes where the table.column value is exactly a certain value.
!=
Select nodes where the table.column value is not a given specific value.
=~
Select nodes where the table.column value matches a given regular expression.
!~
Select nodes where the table.column value does not match a given regular expression.
The nodels command with a specific node and one or more table.attribute parameters is a good substitute for grep’ing through the tab files, as was typically done in xCAT 1.x. This is because nodels will translate any regular expression rows in the tables into their meaning for the specified node. The tab* commands will not do this, instead they will just display the regular expression row verbatim.
-v|--version
Command Version.
-H|--with-fieldname
Force display of table name and column name context for each result
-b|--blame
For values inherited from groups, display which groups provided the inheritence
-S
List all the hidden nodes (FSP/BPA nodes) with other ones.
-?|-h|--help
Display usage message.
To list all defined nodes, enter:
nodelsOutput is similar to:
node1 node2 node3
To list all defined attributes in a table for a node or noderange, enter:
nodels rra001a noderes
Output is similar to:
rra001a: noderes.primarynic: eth0 rra001a: noderes.xcatmaster: rra000 rra001a: noderes.installnic: eth0 rra001a: noderes.netboot: pxe rra001a: noderes.servicenode: rra000 rra001a: noderes.node: rra001a
To list nodes in node group ppc, enter:
nodels ppc
Output is similar to:
ppcnode1 ppcnode2 ppcnode3
To list the groups each node is part of:
nodels all groups
Output is similar to:
node1: groups: all node2: groups: all,storage node3: groups: all,blade
To list the groups each node is part of:
nodels all nodehm.power
Output is similar to:
node1: nodehm.power: blade node2: nodehm.power: ipmi node3: nodehm.power: ipmi
To list the out-of-band mgt method for blade1:
nodels blade1 nodehm.mgt
Output is similar to:
blade1: blade
Listing blades managed through an AMM named ‘amm1’
nodels all mp.mpa==amm1
Output is similar to:
blade1 blade10 blade11 blade12 blade13 blade2 blade3 blade4 blade5 blade6 blade7 blade8 blade9
Listing the switch.switch value for nodes in the second rack:
nodels all nodepos.rack==2 switch.switch
Output is similar to:
n41: switch.switch: switch2 n42: switch.switch: switch2 n43: switch.switch: switch2 n44: switch.switch: switch2 n45: switch.switch: switch2 n46: switch.switch: switch2 n47: switch.switch: switch2 n55: switch.switch: switch2 n56: switch.switch: switch2 n57: switch.switch: switch2 n58: switch.switch: switch2 n59: switch.switch: switch2 n60: switch.switch: switch2
Listing the blade slot number for anything managed through a device with a name beginning with amm:
nodels all mp.mpa=~/^amm.*/ mp.id
Output looks like:
blade1: mp.id: 1 blade10: mp.id: 10 blade11: mp.id: 11 blade12: mp.id: 12 blade13: mp.id: 13 blade2: mp.id: 2 blade3: mp.id: 3 blade4: mp.id: 4 blade5: mp.id: 5 blade6: mp.id: 6 blade7: mp.id: 7 blade8: mp.id: 8 blade9: mp.id: 9
To list the hidden nodes that can’t be seen with other flags. The hidden nodes are FSP/BPAs.
lsdef -S
/opt/xcat/bin/nodels
noderange(3)|noderange.3, tabdump(8)|tabdump.8, lsdef(1)|lsdef.1
nodepurge.1¶
nodepurge - Removes nodes.
The nodepurge automatically removes all nodes from the database and any related configurations used by the node.
After the nodes are removed, the configuration files related to these nodes are automatically updated, including the following files: /etc/hosts, DNS, DHCP. Any kits that are used by the nodes are triggered to automatically update kit configuration and services.
-h|--help
Display usage message.
-v|--version
Command Version
noderange
The nodes to be removed.
To remove nodes compute-000 and compute-001, use the following command:
nodepurge compute-000,compute-001
nodeimport(1)|nodeimport.1, nodechprofile(1)|nodechprofile.1, noderefresh(1)|noderefresh.1, noderange(3)|noderange.3
noderefresh.1¶
noderefresh - Update nodes configurations by running associated kit plugins.
The noderefresh command will update nodes settings, it will call all associated kit plug-in configurations and also services
-h|--help
Display usage message.
-v|--version
Command Version.
noderange
The nodes to be updated.
noderefresh compute-000,compute-001
nodeimport(1)|nodeimport.1, nodechprofile(1)|nodechprofile.1, nodepurge(1)|nodepurge.1, noderange(3)|noderange.3
noderm.1¶
noderm -Removes the nodes in the noderange from all database table.
The noderm command removes the nodes in the input node range.
-h Display usage message.
/opt/xcat/bin/noderm
nodels(1)|nodels.1, nodeadd(8)|nodeadd.8, noderange(3)|noderange.3
nodestat.1¶
nodestat - display the running status of each node in a noderange
nodestat displays and optionally updates the database the running status of a single or range of nodes or groups. See noderange(3)|noderange.3.
- By default, it works as following:
- gets the sshd,pbs_mom,xend port status;
- if none of them are open, it gets the fping status;
- for pingable nodes that are in the middle of deployment, it gets the deployment status;
- for non-pingable nodes, it shows ‘noping’.
When -m is specified and there are settings in the monsetting table, it displays the status of the applications specified in the monsetting table. When -p is specified it shows the power status for the nodes that are not pingable. When -u is specified it saves the status info into the xCAT database. Node’s pingable status and deployment status is saved in the nodelist.status column. Node’s application status is saved in the nodelist.appstatus column.
To specify settings in the monsetting table, use ‘xcatmon’ as the name, ‘apps’ as the key and the value will be a list of comma separated list of application names. For each application, you can specify the port number that can be queried on the nodes to get the running status. Or you can specify a command that can be called to get the node status from. The command can be a command that can be run locally at the management node or the service node for hierarchical cluster, or a command that can be run remotely on the nodes.
The following is an example of the settings in the monsetting table:
name key value
xcatmon apps ssh,ll,gpfs,someapp
xcatmon gpfs cmd=/tmp/mycmd,group=compute,group=service
xcarmon ll port=9616,group=compute
xcatmon someapp dcmd=/tmp/somecmd
Keywords to use:
apps -- a list of comma separated application names whose status will be queried. For how to get the status of each app, look for app name in the key filed in a different row.
port -- the application daemon port number, if not specified, use internal list, then /etc/services.
group -- the name of a node group that needs to get the application status from. If not specified, assume all the nodes in the nodelist table. To specify more than one groups, use group=a,group=b format.
cmd -- the command that will be run locally on mn or sn.
lcmd -- the command that will be run the the mn only.
dcmd -- the command that will be run distributed on the nodes using xdsh <nodes> ....
For commands specified by ‘cmd’ and ‘lcmd’, the input of is a list of comma separated node names, the output must be in the following format:
node1:string1
node2:string2
...
For the command specified by ‘dcmd’, no input is needed, the output can be a string.
-f
Uses fping instead of nmap even if nmap is available. If you seem to be having a problem with false negatives, fping can be more forgiving, but slower.
-m | --usemon
Uses the settings from the monsetting talbe to determine a list of applications that need to get status for.
-p | --powerstat
Gets the power status for the nodes that are ‘noping’.
-u | --updatedb
Updates the status and appstatus columns of the nodelist table with the returned running status from the given nodes.
-v | --version
Print version.
-h | --help
Print help.
- nodestat compute
node1 sshd
node2 sshd
node3 ping
node4 pbs
node5 noping
- nodestat compute -p
node1 sshd
node2 sshd
node3 ping
node4 pbs
node5 noping(Shutting down)
- nodestat compute -u
node1 sshd node2 sshd node3 ping node4 netboot node5 noping
- nodestat compute -m
node1 ping,sshd,ll,gpfs=ok node2 ping,sshd,ll,gpfs=not ok,someapp=something is wrong node3 netboot node4 noping
noderange(3)|noderange.3, nodels(1)|nodels.1, nodeset(8)|nodeset.8
packimage.1¶
packimage - Packs the stateless image from the chroot file system.
packimage [-h| –help]
packimage [-v| –version]
packimage [-o OS] [ -p profile] [-a architecture] [-m method]
packimage imagename
Packs the stateless image from the chroot file system into a file system to be sent to the node for a diskless install. The install dir is setup by using “installdir” attribute set in the site table. The nodetype table “profile” attribute for the node should reflect the profile of the install image.
If no imagename is specified, this command uses the os image definition files from /install/custom/netboot/[os] directory first. If not found, it falls back to the default directory /opt/xcat/share/xcat/netboot/[os]. If a imagename is specified, this command will get all the necessary os image definition files from the osimage and linuximage tables.
imagename specifies the name of a os image definition to be used. The specification for the image is stored in the osimage table and linuximage table.
-h Display usage message.
-v Command Version.
-o Operating system (fedora8, rhel5, sles10,etc)
-p Profile (compute,service)
-a Architecture (ppc64,x86_64,etc)
-m Method (default cpio)
- To pack a fedora8 image for a compute node architecture x86_64 and place it in the /install/netboot/fedora8/x86_64/compute/rootimg.gz file enter:
packimage -o fedora8 -p compute -a x86_64
This would use the package information from the /install/custom/netboot/fedora/compute* files first. If not found it uses /opt/xcat/share/xcat/netboot/fedora/compute* files.
/opt/xcat/sbin/packimage
This command is part of the xCAT software product.
genimage(1)|genimage.1
pasu.1¶
pasu - run the ASU to many nodes in parallel
pasu [-V] [-d] [-n] [-l user] [-p passwd] [-f fanout] [-i hostname-suffix] noderange command
pasu [-V] [-d] [-n] [-l user] [-p passwd] [-f fanout] [-i hostname-suffix] -b batchfile noderange
pasu [-h | --help]
The pasu command runs the ASU command in out-of-band mode in parallel to multiple nodes. Out-of-band mode means that ASU connects from the xCAT management node to the IMM (BMC) of each node to set or query the ASU settings. To see all of the ASU settings available on the node, use the “show all” command. To query or set multiple values, use the -b (batch) option. To group similar output from multiple nodes, use xcoll(1)|xcoll.1.
Before running pasu, you must install the ASU RPM from IBM. You can download it from the IBM Fix Central site. You also must configure the IMMs properly according to xCAT documentation. Run “rpower noderange stat” to confirm that the IMMs are configured properly.
-n|--nonodecheck
Do not send the noderange to xcatd to expand it into a list of nodes. Use the noderange exactly as it is specified to pasu. In this case, the noderange must be a simple list of comma-separated hostnames of the IMMs.
-l|--loginname username
The username to use to connect to the IMMs. If not specified, the row in the xCAT passwd table with key “ipmi” will be used to get the username.
-p|--passwd passwd
The password to use to connect to the IMMs. If not specified, the row in the xCAT passwd table with key “ipmi” will be used to get the password.
-f|--fanout
How many processes to run in parallel simultaneously. The default is 64. You can also set the XCATPSHFANOUT environment variable.
-b|--batch -batchfile
A simple text file that contains multiple ASU commands, each on its own line.
-d|--donotfilter
By default, pasu filters out (i.e. does not display) the standard initial output from ASU:
IBM Advanced Settings Utility version 9.30.79N Licensed Materials - Property of IBM (C) Copyright IBM Corp. 2007-2012 All Rights Reserved Connected to IMM at IP address node2-immIf you want this output to be displayed, use this flag.
-i|--interface hostname-suffix
The hostname suffix to be appended to the node names.
-V|--verbose
Display verbose messages.
-h|--help
Display usage message.
To display the Com1ActiveAfterBoot setting on 2 nodes:
pasu node1,node2 show DevicesandIOPorts.Com1ActiveAfterBoot
Output is similar to:
node1: DevicesandIOPorts.Com1ActiveAfterBoot=Enable node2: DevicesandIOPorts.Com1ActiveAfterBoot=Enable
To display the Com1ActiveAfterBoot setting on all compute nodes:
pasu compute show DevicesandIOPorts.Com1ActiveAfterBoot | xcoll
Output is similar to:
==================================== compute ==================================== DevicesandIOPorts.Com1ActiveAfterBoot=Enable
To set several settings on all compute nodes, create a batch file called (for example) asu-settings with contents:
set DevicesandIOPorts.Com1ActiveAfterBoot Enable set DevicesandIOPorts.SerialPortSharing Enable set DevicesandIOPorts.SerialPortAccessMode Dedicated set DevicesandIOPorts.RemoteConsole Enable
Then run:
pasu -b asu-settings compute | xcoll
Output is similar to:
==================================== compute ==================================== Batch mode start. [set DevicesandIOPorts.Com1ActiveAfterBoot Enable] DevicesandIOPorts.Com1ActiveAfterBoot=Enable [set DevicesandIOPorts.SerialPortSharing Enable] DevicesandIOPorts.SerialPortSharing=Enable [set DevicesandIOPorts.SerialPortAccessMode Dedicated] DevicesandIOPorts.SerialPortAccessMode=Dedicated [set DevicesandIOPorts.RemoteConsole Enable] DevicesandIOPorts.RemoteConsole=Enable Beginning intermediate batch update. Waiting for command completion status. Command completed successfully. Completed intermediate batch update. Batch mode competed successfully.
To confirm that all the settings were made on all compute nodes, create a batch file called (for example) asu-show with contents:
show DevicesandIOPorts.Com1ActiveAfterBoot show DevicesandIOPorts.SerialPortSharing show DevicesandIOPorts.SerialPortAccessMode show DevicesandIOPorts.RemoteConsole
Then run:
pasu -b asu-show compute | xcoll
Output is similar to:
==================================== compute ==================================== Batch mode start. [show DevicesandIOPorts.Com1ActiveAfterBoot] DevicesandIOPorts.Com1ActiveAfterBoot=Enable [show DevicesandIOPorts.SerialPortSharing] DevicesandIOPorts.SerialPortSharing=Enable [show DevicesandIOPorts.SerialPortAccessMode] DevicesandIOPorts.SerialPortAccessMode=Dedicated [show DevicesandIOPorts.RemoteConsole] DevicesandIOPorts.RemoteConsole=Enable Batch mode competed successfully.
/opt/xcat/bin/pasu
noderange(3)|noderange.3, rpower(1)|rpower.1, xcoll(1)|xcoll.1
pcons.1¶
Runs the a command to the noderange using the console.
pcons 1,3 stat
pcons all,-129-256 stat
psh(1)|psh.1
pgsqlsetup.1¶
pgsqlsetup - Sets up the PostgreSQL database for xCAT to use.
pgsqlsetup {-h | --help}
pgsqlsetup {-v | --version}
pgsqlsetup {-i | --init} [-N|nostart] [-P|-**-PCM] [-o|--setupODBC] [**-V | --verbose]
pgsqlsetup - Sets up the PostgreSQL database for xCAT to use. The pgsqlsetup script is run on the Management Node as root after the PostgreSQL code has been installed. The xcatd daemon will be stopped during migration. No xCAT commands should be run during the init process, because we will be migrating the xCAT database to PostgreSQL and restarting the xcatd daemon as well as the PostgreSQL daemon. For full information on all the steps that will be done reference One password must be supplied for the setup, a password for the xcatadm unix id and the same password for the xcatadm database id. The password will be prompted for interactively or you can set the XCATPGPW environment variable to the password and then there will be no prompt.
-h|--help
Displays the usage message.
-v|--version
Displays the release version of the code.
-V|--verbose
Displays verbose messages.
-i|--init
The init option is used to setup an installed PostgreSQL database so that xCAT can use the database. This involves creating the xcat database, the xcat admin id, allowing access to the xcatdb database by the Management Node. It customizes the postgresql.conf configuration file, adds the management server to the pg_hba.conf and starts the PostgreSQL server. It also backs up the current xCAT database and restores it into the newly setup xcatdb PostgreSQL database. It creates the /etc/xcat/cfgloc file to point the xcatd daemon to the PostgreSQL database and restarts the xcatd daemon using the database. On AIX, it additionally setup the xcatadm unix id and the postgres id and group. For AIX, you should be using the PostgreSQL rpms available from the xCAT website. For Linux, you should use the PostgreSQL rpms shipped with the OS. You can chose the -o option, to run after the init. To add additional nodes to access the PostgreSQL server, setup on the Management Node, edit the pg_hba.conf file.
For more documentation see:Setting_Up_PostgreSQL_as_the_xCAT_DB
-N|--nostart
This option with the -i flag will create the database, but will not backup and restore xCAT tables into the database. It will create the cfgloc file such that the next start of xcatd will try and contact the database. This can be used to setup the xCAT PostgreSQL database during or before install.
-P|--PCM
This option sets up PostgreSQL database to be used with xCAT running with PCM.
-o|--odbc
This option sets up the ODBC /etc/../odbcinst.ini, /etc/../odbc.ini and the .odbc.ini file in roots home directory will be created and initialized to run off the xcatdb PostgreSQL database.
XCATPGPW
The password to be used to setup the xCAT admin id for the database.
*
To setup PostgreSQL for xCAT to run on the PostgreSQL xcatdb database :
pgsqlsetup -i
*
To setup the ODBC for PostgreSQL xcatdb database access :
pgsqlsetup -o
pping.1¶
pping [-i | --interface interfaces] [-f | --use_fping] noderange
pping [-h | --help]
pping {-v | --version}
pping is a utility used to ping a list of nodes in parallel. pping will return an unsorted list of nodes with a ping or noping status. pping front-ends nmap or fping if available.
This command does not support the xcatd client/server communication. It must be run on the management node.
-i | --interface interfaces
A comma separated list of network interface names that should be pinged instead of the interface represented by the nodename/hostname. The following name resolution convention is assumed: an interface is reachable by the hostname <nodename>-<interface>. For example, the ib2 interface on node3 has a hostname of node3-ib2.
If more than one interface is specified, each interface will be combined with the nodenames as described above and will be pinged in turn.
-f | --use_fping
Use fping instead of nmap
-h | --help
Show usage information.
-v | --version
Display the installed version of xCAT.
pping all
node1: ping node2: ping node3: noping
pping all -i ib0,ib1
node1-ib0: ping node2-ib0: ping node3-ib0: noping node1-ib1: ping node2-ib1: ping node3-ib1: noping
psh(1)|psh.1, noderange(3)|noderange.3
ppping.1¶
ppping [-i | --interface interfaces] [-d | --debug] [-V | --verbose] [-q | --quiet] [-s | --serial] noderange
ppping [-h | --help]
pping {-v | --version}
ppping is a utility used to test the connectivity between nodes in the noderange using ping. By default, ppping will return an unsorted list of the node pairs that are not able to ping each other, or a message that all nodes are pingable. More or less output can be controlled by the -V and -q options. ppping front-ends pping and xdsh.
-s
Ping serially instead of in parallel.
-i | --interface interfaces
A comma separated list of network interface names that should be pinged instead of the interface represented by the nodename/hostname. The following name resolution convention is assumed: an interface is reachable by the hostname <nodename>-<interface>. For example, the ib2 interface on node3 has a hostname of node3-ib2.
If more than one interface is specified, each interface will be combined with the nodenames as described above and will be pinged in turn.
-V | --verbose
Display verbose output. The result of every ping attempt from every node will be displayed. Without this option, just a summary of the successful pings are displayed, along with all of the unsuccessful pings.
-q | --quiet
Display minimum output: just the unsuccessful pings. This option has the effect that if all pings are successful, nothing is displayed. But it also has the performance benefit that each node does not have to send successful ping info back to the management node.
-d | --debug
Print debug information.
-h | --help
Show usage information.
-v | --version
Display the installed version of xCAT.
ppping all -q
blade7: node2: noping blade8: node2: noping blade9: node2: noping devmaster: node2: noping node2: noping
ppping node1,node2 -i ib0,ib1,ib2,ib3
node1: pinged all nodes successfully on interface ib0 node1: pinged all nodes successfully on interface ib1 node1: pinged all nodes successfully on interface ib2 node1: pinged all nodes successfully on interface ib3 node2: pinged all nodes successfully on interface ib0 node2: pinged all nodes successfully on interface ib1 node2: pinged all nodes successfully on interface ib2 node2: pinged all nodes successfully on interface ib3
psh(1)|psh.1, pping(1)|pping.1
prsync.1¶
prsync - parallel rsync
prsync filename [filename ...] noderange:destinationdirectory
prsync [-o rsync options] [-f fanout] [filename filename ...] [directory directory ...] noderange:destinationdirectory
prsync {-h | --help | -v | --version}
prsync is a front-end to rsync for a single or range of nodes and/or groups in parallel.
Note: this command does not support the xcatd client/server communication and therefore must be run on the management node. It does not support hierarchy, use xdcp -F to run rsync from the management node to the compute node via a service node
prsync is NOT multicast, but is parallel unicasts.
rsyncopts
rsync options. See rsync(1).
-f fanout
Specifies a fanout value for the maximum number of concur- rently executing remote shell processes.
filename
A space delimited list of files to rsync.
directory
A space delimited list of directories to rsync.
noderange:destination
A noderange(3)|noderange.3 and destination directory. The : is required.
-h | --help
Print help.
-v | --version
Print version.
XCATPSHFANOUT
Specifies the fanout value. This variable is overridden by the -f flag. Default is 64.
noderange(3)|noderange.3, pscp(1)|pscp.1, pping(1)|pping.1, psh(1)|psh.1
pscp.1¶
pscp - parallel remote copy
pscp [-i suffix] [scp options ...] [-f fanout] filename [filename ...] noderange:destinationdirectory
pscp {-h | --help | -v | --version}
pscp is a utility used to copy a single or multiple set of files and/or directories to a single or range of nodes and/or groups in parallel.
pscp is a front-end to the remote copy scp.
Note: this command does not support the xcatd client/server communication and therefore must be run on the management node. It does not support hierarchy, use xdcp to run remote copy command from the management node to the compute node via a service node.
pscp is NOT multicast, but is parallel unicasts.
-f fanout
Specifies a fanout value for the maximum number of concur- rently executing remote shell processes.
-i suffix
Interfaces to be used.
scp options
See scp(1)
filename
A space delimited list of files to copy. If -r is passed as an scp option, directories may be specified as well.
noderange:destination
A noderange(3)|noderange.3 and destination directory. The : is required.
-h | --help
Print help.
-v | --version
Print version.
XCATPSHFANOUT
Specifies the fanout value. This variable is overridden by the -f flag. Default is 64.
pscp -r /usr/local node1,node3:/usr/localpscp passwd group rack01:/etc
noderange(3)|noderange.3, pping(1)|pping.1, prsync(1)|prsync.1, psh(1)|psh.1
psh.1¶
psh - parallel remote shell
psh [-i interface] [-f fanout] [-l user] noderange command
psh {-h | --help | -v | --version}
psh is a utility used to run a command across a list of nodes in parallel.
ssh must be set up to allow no prompting for psh to work.
Note:
This command does not run through xcatd like most xCAT commands do. This means you must either run it on the management node, or have a network connection between your machine and the nodes. It does not support hierarchy, use xdsh to run remote command from the management node to the compute node via a service node.
psh arguments need to precede noderange, otherwise, you will get unexpected errors.
-i interface
The NIC on the node that psh should communicate with. For example, if interface is eth1, then psh will concatenate -eth1 to the end of every node name before ssh’ing to it. This assumes those host names have been set up to resolve to the IP address of each of the eth1 NICs.
-f fanout
Specifies a fanout value for the maximum number of concur- rently executing remote shell processes.
-l user
Log into the nodes as the specified username. The default is to use the same username as you are running the psh command as.
-n|--nonodecheck
Do not send the noderange to xcatd to expand it into a list of nodes. Instead, use the noderange exactly as it is specified. In this case, the noderange must be a simple list of comma-separated hostnames of the nodes. This allows you to run psh even when xcatd is not running.
noderange
See noderange(3)|noderange.3.
command
Command to be run in parallel. If no command is give then pshenters interactive mode. In interactive mode a “>” prompt is displayed. Any command entered is executed in parallel to the nodes in the noderange. Use “exit” or “Ctrl-D” to end the interactive session.
-h | --help
Print help.
XCATPSHFANOUT
Specifies the fanout value. This variable is overridden by the -f flag. Default is 64.
*
Run uptime on 3 nodes:
psh node4-node6 uptime
node4: Sun Aug 5 17:42:06 MDT 2001 node5: Sun Aug 5 17:42:06 MDT 2001 node6: Sun Aug 5 17:42:06 MDT 2001
*
Run a command on some BladeCenter management modules:
psh amm1-amm5 ‘info -T mm[1]’
*
Remove the tmp files on the nodes in the 1st frame:
psh rack01 ‘rm -f /tmp/\‘*
Notice the use of ‘’ to forward shell expansion. This is not necessary in interactive mode.
noderange(3)|noderange.3, pscp(1)|pscp.1, pping(1)|pping.1, prsync(1)|prsync.1
pushinitrd.1¶
pushinitrd - queries your SoftLayer account and gets attributes for each server.
The pushinitrd command copies the initrd, kernel, params, and static IP info to nodes, so they can be net installed even across vlans (w/o setting up pxe/dhcp broadcast relay). This assumes a working OS is on the nodes. Before running this command, you must run nodeset for these nodes. All of the nodes given to one invocation of pushinitrd must be using the same osimage.
Before using this command, if will be most convenient if you exchange the ssh keys using:
xdsh <noderange> -K
-w waittime
The number of seconds the initrd should wait before trying to communicate over the network. The default is 75. This translates into the netwait kernel parameter and is usually needed in a SoftLayer environment because it can take a while for a NIC to be active after changing state.
-?|-h|--help
Display usage message.
-v|--version
Command Version.
Configure nodes for net installing in a SoftLayer environment:
pushinitrd <noderange>
/opt/xcat/bin/pushinitrd
getslnodes(1)|getslnodes.1
rbeacon.1¶
rbeacon noderange {on | blink | off | stat}
rbeacon [-h | --help]
rbeacon {-v | --version}
rbeacon Turns beacon (a light on the front of the physical server) on/off/blink or gives status of a node or noderange.
rbeacon 1,3 off
rbeacon 14-56,70-203 on
rbeacon 1,3,14-56,70-203 blink
rbeacon all,-129-256 stat
noderange(3)|noderange.3, rpower(1)|rpower.1
rbootseq.1¶
rbootseq [-h | --help | -v | --version]
rbootseq noderange {hd0 | hd1 | hd2 | hd3 | net | iscsi | iscsicrit | cdrom | usbflash | floppy | none | list | stat},...
rbootseq noderange {hd | net1 | net2 | net3 | net4 | cdrom | usbflash | floppy | none},...
rbootseq noderange [hfi|net]
For Blade specific:
rbootseq sets the boot sequence (the order in which boot devices should be tried) for the specified blades. Up to four different medium/devices can be listed, separated by commas. The boot sequence will remain in effect for these blades until set differently.
For PPC (using Direct FSP Management) specific:
rbootseq sets the ethernet (net) or hfi device as the first boot device for the specified PPC LPARs. The rbootseq command requires that the ethernet or hfi mac address is stored in the mac table, and that the network information is correct in the networks table.
hd0 | harddisk0 | hd | harddisk
The first hard disk.
hd1 | harddisk1
The second hard disk.
hd2 | harddisk2
The third hard disk.
hd3 | harddisk3
The fourth hard disk.
n | net | network
Boot over the ethernet network, using a PXE or BOOTP broadcast.
n | net | network | net1 | nic1 (HP Blade Only)
Boot over the first ethernet network, using a PXE or BOOTP broadcast.
net2 | nic2 (HP Blade Only)
Boot over the second ethernet network, using a PXE or BOOTP broadcast.
net3 | nic3 (HP Blade Only)
Boot over the third ethernet network, using a PXE or BOOTP broadcast.
net3 | nic3 (HP Blade Only)
Boot over the fourth ethernet network, using a PXE or BOOTP broadcast.
hfi
Boot p775 nodes over the HFI network, using BOOTP broadcast.
iscsi
Boot to an iSCSI disk over the network.
iscsicrit
??
cd | cdrom
The CD or DVD drive.
usbflash | usb | flash
A USB flash drive.
floppy
The floppy drive.
none
If it gets to this part of the sequence, do not boot. Can not be specified 1st, or before any real boot devices.
list | stat
Display the current boot sequence.
Set blades 14-56 and 70-203 to try to boot first from the CD drive, then the floppy drive, then the network, and finally from the 1st hard disk:
rbootseq blade[14-56],blade[70-203] c,f,n,hd0
rsetboot(1)|rsetboot.1
rcons.1¶
rcons - remotely accesses the serial console of a node
rcons provides access to a single remote node serial console, using the out-of-band infrastructure for the node (e.g. BMC, Management Module, HMC, KVM, etc.). It uses the conserver open source package to provide one read-write and multiple read-only instances of the console, plus console logging.
If conserver-host is specified, the conserver daemon on that host will be contacted, instead of on the local host.
To exit the console session, enter: <ctrl><shift>e c .
-f
If another console for this node is already open in read-write mode, force that console into read-only (spy) mode, and open this console in read-write mode. If -f is not specified, this console will be put in spy mode if another console is already open in read-write mode. The -f flag can not be used with the -s flag.
-s
Open the console in read-only (spy) mode, in this mode all the escape sequences work, but all other keyboard input is discarded. The -s flag can not be used with the -f flag.
-h | --help
Print help.
-v | --version
Print version.
nodehm table - xCAT node hardware management table. See nodehm(5)|nodehm.5 for further details. This is used to determine the console access method.
rcons node5
wcons(1)|wcons.1
regnotif.1¶
regnotif - Registers a Perl module or a command that will get called when changes occur in the desired xCAT database tables.
regnotif [-h| –help]
regnotif [-v| –version]
regnotif *filename tablename*[,tablename]... [-o|–operation actions]
This command is used to register a Perl module or a command to the xCAT notification table. Once registered, the module or the command will get called when changes occur in the xCAT database tables indicated by tablename. The changes can be row addition, deletion and update which are specified by actions.
filename is the path name of the Perl module or command to be registered. tablename is the name of the table that the user is interested in.
-h | -help Display usage message.
**-v | -version ** Command Version.
-V | -verbose Verbose output.
-o | -operation specifies the database table actions that the user is interested in. It is a comma separated list. ‘a’ for row addition, ‘d’ for row deletion and ‘u’ for row update.
- To register a Perl module that gets invoked when rows get added or deleted. in the nodelist and the nodehm tables, enter:
regnotif /opt/xcat/lib/perl/xCAT_monitoring/mycode.pm nodelist,nodhm -o a,d
- To register a command that gets invoked when rows get updated in the switch table, enter:
regnotif /usr/bin/mycmd switch -o u
/opt/xcat/bin/regnotif
unregnotif(1)|unregnotif.1
renergy.1¶
renergy - remote energy management tool
Power 6 server specific :
renergy noderange [-V] { all | [savingstatus] [cappingstatus] [cappingmaxmin] [cappingvalue] [cappingsoftmin] [averageAC] [averageDC] [ambienttemp] [exhausttemp] [CPUspeed] [syssbpower] [sysIPLtime]}
renergy noderange [-V] { savingstatus={on | off} | cappingstatus={on | off} | cappingwatt=watt | cappingperc=percentage }
Power 7 server specific :
renergy noderange [-V] { all | [savingstatus] [dsavingstatus] [cappingstatus] [cappingmaxmin] [cappingvalue] [cappingsoftmin] [averageAC] [averageDC] [ambienttemp] [exhausttemp] [CPUspeed] [syssbpower] [sysIPLtime] [fsavingstatus] [ffoMin] [ffoVmin] [ffoTurbo] [ffoNorm] [ffovalue]}
renergy noderange [-V] { savingstatus={on | off} | dsavingstatus={on-norm | on-maxp | off} | fsavingstatus={on | off} | ffovalue=MHZ | cappingstatus={on | off} | cappingwatt=watt | cappingperc=percentage }
Power 8 server specific :
renergy noderange [-V] { all | [savingstatus] [dsavingstatus] [averageAC] [averageAChistory] [averageDC] [averageDChistory] [ambienttemp] [ambienttemphistory] [exhausttemp] [exhausttemphistory] [fanspeed] [fanspeedhistory] [CPUspeed] [CPUspeedhistory] [syssbpower] [sysIPLtime] [fsavingstatus] [ffoMin] [ffoVmin] [ffoTurbo] [ffoNorm] [ffovalue]}
renergy noderange [-V] { savingstatus={on | off} | dsavingstatus={on-norm | on-maxp | off} | fsavingstatus={on | off} | ffovalue=MHZ }
NOTE: The setting operation for Power 8 server is only supported for the server which is running in PowerVM mode. Do NOT run the setting for the server which is running in OPAL mode.
BladeCenter specific :
For Management Modules:
renergy noderange [-V] { all | pd1all | pd2all | [pd1status] [pd2status] [pd1policy] [pd2policy] [pd1powermodule1] [pd1powermodule2] [pd2powermodule1] [pd2powermodule2] [pd1avaiablepower] [pd2avaiablepower] [pd1reservedpower] [pd2reservedpower] [pd1remainpower] [pd2remainpower] [pd1inusedpower] [pd2inusedpower] [availableDC] [averageAC] [thermaloutput] [ambienttemp] [mmtemp] }
For a blade server nodes:
renergy noderange [-V] { all | [averageDC] [capability] [cappingvalue] [CPUspeed] [maxCPUspeed] [savingstatus] [dsavingstatus] }
renergy noderange [-V] { savingstatus={on | off} | dsavingstatus={on-norm | on-maxp | off} }
Flex specific :
For Flex Management Modules:
renergy noderange [-V] { all | [powerstatus] [powerpolicy] [powermodule] [avaiablepower] [reservedpower] [remainpower] [inusedpower] [availableDC] [averageAC] [thermaloutput] [ambienttemp] [mmtemp] }
For Flex node (power and x86):
renergy noderange [-V] { all | [averageDC] [capability] [cappingvalue] [cappingmaxmin] [cappingmax] [cappingmin] [cappingGmin] [CPUspeed] [maxCPUspeed] [savingstatus] [dsavingstatus] }
renergy noderange [-V] { cappingstatus={on | off} | cappingwatt=watt | cappingperc=percentage | savingstatus={on | off} | dsavingstatus={on-norm | on-maxp | off} }
iDataPlex specific :
renergy noderange [-V] [ { cappingmaxmin | cappingmax | cappingmin } ] [cappingstatus] [cappingvalue] [relhistogram]
renergy noderange [-V] { cappingstatus={on | enable | off | disable} | {cappingwatt|cappingvalue}=watt }
This renergy command can be used to manage the energy consumption of IBM servers which support IBM EnergyScale technology. Through this command, user can query and set the power saving and power capping status, and also can query the average consumed energy, the ambient and exhaust temperature, the processor frequency for a server.
renergy command supports IBM POWER6, POWER7 and POWER8 rack-mounted servers, BladeCenter management modules, blade servers, and iDataPlex servers. For Power6 and Power7 rack-mounted servers, the following specific hardware types are supported: 8203-E4A, 8204-E8A, 9125-F2A, 8233-E8B, 8236-E8C. For Power8 server, there’s no hardware type restriction.
The parameter noderange needs to be specified for the renergy command to get the target servers. The noderange should be a list of CEC node names, blade management module node names or blade server node names. Lpar name is not acceptable here.
renergy command can accept multiple of energy attributes to query or one of energy attribute to set. If only the attribute name is specified, without the ‘=’, renergygets and displays the current value. Otherwise, if specifying the attribute with ‘=’ like ‘savingstatus=on’, renergy will set the attribute savingstatus to value ‘on’.
The attributes listed in the SYNOPSIS section are which ones can be handled by renergy command. But for each specific type of server, there are some attributes that are not supported. If user specifies an attribute which is not supported by a specific server, the return value of this attribute will be ‘na’.
The supported attributes for each specific system p hardware type is listed as follows:
8203-E4A, 8204-E8A
Supported attributes:
Query: savingstatus,cappingstatus,cappingmin,cappingmax, cappingvalue,cappingsoftmin,averageAC,averageDC,ambienttemp, exhausttemp,CPUspeed,syssbpower,sysIPLtime
Set: savingstatus,cappingstatus,cappingwatt,cappingperc
9125-F2A
Supported attributes:
Query: savingstatus,averageAC,ambienttemp,exhausttemp, CPUspeed
Set: savingstatus
8233-E8B, 8236-E8C
Supported attributes:
Query: savingstatus,dsavingstatus,cappingstatus,cappingmin, cappingmax,cappingvalue,cappingsoftmin,averageAC,averageDC, ambienttemp,exhausttemp,CPUspeed,syssbpower,sysIPLtime
Set: savingstatus,dsavingstatus,cappingstatus,cappingwatt, cappingperc
9125-F2C, 9119-FHB
Supported attributes:
Query: savingstatus,dsavingstatus,cappingstatus,cappingmin, cappingmax,cappingvalue,cappingsoftmin,averageAC,averageDC, ambienttemp,exhausttemp,CPUspeed,syssbpower,sysIPLtime, fsavingstatus,ffoMin,ffoVmin,ffoTurbo,ffoNorm,ffovalue
Set: savingstatus,dsavingstatus,cappingstatus,cappingwatt, cappingperc,fsavingstatus,ffovalue
Non of Above
For the machine type which is not in the above list, the following attributes can be tried but not guaranteed:
Query: savingstatus,dsavingstatus,cappingstatus,cappingmin, cappingmax,,cappingvalue,cappingsoftmin,averageAC,averageDC, ambienttemp,exhausttemp,CPUspeed,syssbpower,sysIPLtime
Set: savingstatus,dsavingstatus,cappingstatus,cappingwatt, cappingperc
Note: For system P CEC nodes, each query operation for attribute CPUspeed, averageAC or averageDC needs about 30 seconds to complete. The query for others attributes will get response immediately.
For the Power6 and Power7 nodes, the renergy command depends on the Energy Management Plugin xCAT-pEnergy to communicate with server. xCAT-pEnergy can be downloaded from the IBM web site: http://www.ibm.com/support/fixcentral/. (Other Software -> EM)
NOTE: Power8 nodes don’t need this specific energy management package.
For iDataPlex nodes, the renergy command depends on the Energy Management Plugin xCAT-xEnergy to communicate with server. This plugin must be requested from IBM.
(The support for BladeCenter energy management is built into base xCAT, so no additional plugins are needed for BladeCenter.)
-h | --help
Display the usage message.
-v | --version
Display the version information.
-V
Verbose output.
all
Query all energy attributes which supported by the specific type of hardware.
For Power8 machines, will not display the attributes for historical records.
pd1all
Query all energy attributes of the power domain 1 for blade management module node.
pd2all
Query all energy attributes of the power domain 2 for blade management module node.
ambienttemp
Query the current ambient temperature. (Unit is centigrade)
ambienttemphistory
Query the historical records which were generated in last one hour for ambienttemp.
availableDC
Query the total DC power available for the entire blade center chassis.
averageAC
Query the average power consumed (Input). (Unit is watt)
Note: For 9125-F2A,9125-F2C server, the value of attribute averageAC is the aggregate for all of the servers in a rack.
Note: For Blade Center, the value of attribute averageAC is the total AC power being consumed by all modules
in the chassis. It also includes power consumed by the ChassisCooling Devices for BCH chassis.
averageAChistory
Query the historical records which were generated in last one hour for averageAC.
averageDC
Query the average power consumed (Output). (Unit is watt)
averageDChistory
Query the historical records which were generated in last one hour for averageDC.
capability
Query the Power Capabilities of the blade server.
staticPowerManagement: the module with the static worst case power values.
fixedPowermanagement: the module with the static power values but ability to throttle.
dynamicPowerManagement: the module with power meter capability, measurement enabled, but capping disabled.
dynamicPowerMeasurement1: the module with power meter capability, measurement enabled, phase 1 only
dynamicPowerMeasurement2: the module with power meter capability, measurement enabled, phase 2 or higher
dynamicPowerMeasurementWithPowerCapping: the module with power meter capability, both measurement and capping enabled, phase 2 or higher
cappingGmin
Query the Guaranteed Minimum power capping value in watts.
cappingmax
Query the Maximum of power capping value in watts.
cappingmaxmin
Query the Maximum and Minimum of power capping value in watts.
cappingmin
Query the Minimum of power capping value in watts.
cappingperc=percentage
Set the power capping value base on the percentage of the max-min of capping value which getting from cappingmaxmim attribute. The valid value must be from 0 to 100.
cappingsoftmin
Query the minimum value that can be assigned to power capping without guaranteed enforceability. (Unit is watt)
cappingstatus
Query the power capping status. The result should be ‘on’ or ‘off’.
cappingstatus={on | off}
Set the power capping status. The value must be ‘on’ or ‘off’. This is the switch to turn on or turn off the power capping function.
cappingvalue
Query the current power capping value. (Unit is watt)
cappingwatt=watt
Set the power capping value base on the watt unit.
If the ‘watt’ > maximum of cappingmaxmin or ‘watt’ < cappingsoftmin, the setting operation will be failed. If the ‘watt’ > cappingsoftmin and ‘watt’ < minimum of cappingmaxmin, the value can NOT be guaranteed.
CPUspeed
Query the effective processor frequency. (Unit is MHz)
CPUspeedhistory
Query the historical records which were generated in last one hour for CPUspeed
dsavingstatus
Query the dynamic power saving status. The result should be ‘on-norm’, ‘on-maxp’ or ‘off’.
If turning on the dynamic power saving, the processor frequency and voltage will be dropped dynamically based on the core utilization. It supports two modes for turn on state:
on-norm - means normal, the processor frequency cannot exceed the nominal value;
on-maxp - means maximum performance, the processor frequency can exceed the nominal value.
dsavingstatus={on-norm | on-maxp | off}
Set the dynamic power saving. The value must be ‘on-norm’, ‘on-maxp’ or ‘off’.
The dsavingstatus setting operation needs about 2 minutes to take effect. (The used time depends on the hardware type)
The dsavingstatus only can be turned on when the savingstatus is in turn off status.
exhausttemp
Query the current exhaust temperature. (Unit is centigrade)
exhausttemphistory
Query the historical records which were generated in last one hour for exhausttemp
fanspeed
Query the fan speed for all the fans which installed in this node. (Unit is RPM - Rotations Per Minute))
If there are multiple fans for a node, multiple lines will be output. And a fan name in bracket will be appended after fanspped attribute name.
fanspeedhistory
Query the historical records which were generated in last one hour for fanspeed.
ffoMin
Query the minimum cpu frequency which can be set for FFO. (Fixed Frequency Override)
ffoNorm
Query the maximum cpu frequency which can be set for FFO.
ffoTurbo
Query the advertised maximum cpu frequency (selling point).
ffoVmin
Query the minimum cpu frequency which can be set for dropping down the voltage to save power. That means when you drop the cpu frequency from the ffoVmin to ffoVmin, the voltage won’t change, then there’s no obvious power to be saved.
ffovalue
Query the current value of FFO.
ffovalue=MHZ
Set the current value of FFO. The valid value of ffovalue should be between the ffoMin and ffoNorm.
Note1: Due to the limitation of firmware, the frequency in the range 3501 MHz - 3807 MHz can NOT be set to ffovalue. This range may be changed in future.
Note2: The setting will take effect only when the fsavingstatus is in ‘on’ status. But you need to set the ffovalue to a valid value before enabling the fsavingstatus. (It’s a limitation of the initial firmware and will be fixed in future.)
The ffovalue setting operation needs about 1 minute to take effect.
fsavingstatus
Query the status of FFO. The result should be ‘on’ or ‘off’. ‘on’ - enable; ‘off’ - disable.
fsavingstatus={on | off}
Set the status of FFO. The value must be ‘on’ or ‘off’.
‘on’ - enable. It will take effect only when the ffovaluehas been set to a valid value.
‘off’ -disable. It will take effect immediately.
Note: See the Note2 of ffovalue=MHZ.
maxCPUspeed
Query the maximum processor frequency. (Unit is MHz)
mmtemp
Query the current temperature of management module. (Unit is centigrade)
pd1status | powerstatus
Query the status of power domain 1 for blade management module node.
Note: for the attribute without the leading ‘pd1’ which means there’s only one power doamin in the chassis.
pd1policy | powerpolicy
Query the power management policy of power domain 1.
pd1powermodule1 | powermodule
Query the First Power Module capacity in power domain 1.
pd1powermodule2 | powermodule
Query the Second Power Module capacity in power domain 1.
pd1avaiablepower | avaiablepower
Query the total available power in power domain 1.
pd1reservedpower | reservedpower
Query the power that has been reserved for power domain 1.
pd1remainpower | remainpower
Query the remaining power available in power domain 1.
pd1inusedpower | inusedpower
Query the total power being used in power domain 1.
pd2status
Query the status of power domain 2 for blade management module node.
pd2policy
Query the power management policy of power domain 2.
pd2powermodule1
Query the First Power Module capacity in power domain 2.
pd2powermodule2
Query the Second Power Module capacity in power domain 2.
pd2avaiablepower
Query the total available power in power domain 2.
pd2reservedpower
Query the power that has been reserved for power domain 2.
pd2remainpower
Query the remaining power available in power domain 2.
pd2inusedpower
Query the total power being used in power domain 2.
relhistogram
Query histogram data for wattage information
savingstatus
Query the static power saving status. The result should be ‘on’ or ‘off’. ‘on’ - enable; ‘off’ - disable.
savingstatus={on | off}
Set the static power saving. The value must be ‘on’ or ‘off’.
If turning on the static power saving, the processor frequency and voltage will be dropped to a fixed value to save energy.
The savingstatus setting operation needs about 2 minutes to take effect. (The used time depends on the hardware type)
The savingstatus only can be turned on when the dsavingstatus is in turn off status.
sysIPLtime
Query the time used from FSP standby to OS standby. (Unit is Second)
syssbpower
Query the system power consumed prior to power on. (Unit is Watt)
thermaloutput
Query the thermal output (load) in BTUs per hour for the blade center chassis.
1
Query all attributes which CEC1,CEC2 supported.
renergy CEC1,CEC2 all
The output of the query operation:
CEC1: savingstatus: off CEC1: dsavingstatus: off CEC1: cappingstatus: off CEC1: cappingmin: 1953 W CEC1: cappingmax: 2358 W CEC1: cappingvalue: 2000 W CEC1: cappingsoftmin: 304 W CEC1: averageAC: na CEC1: averageDC: na CEC1: ambienttemp: na CEC1: exhausttemp: na CEC1: CPUspeed: na CEC1: syssbpower: 40 W CEC1: sysIPLtime: 900 S CEC2: savingstatus: off CEC2: cappingstatus: off CEC2: cappingmin: 955 W CEC2: cappingmax: 1093 W CEC2: cappingvalue: 1000 W CEC2: cappingsoftmin: 226 W CEC2: averageAC: 627 W CEC2: averageDC: 531 W CEC2: ambienttemp: 25 C CEC2: exhausttemp: 40 C CEC2: CPUspeed: 4695 MHz
2
Query the fanspeed attribute for Power8 CEC.
renergy CEC1 fanspeed
The output of the query operation:
CEC1: fanspeed (Fan U78CB.001.WZS00MA-A1 00002101): 5947 RPM CEC1: fanspeed (Fan U78CB.001.WZS00MA-A2 00002103): 6081 RPM CEC1: fanspeed (Fan U78CB.001.WZS00MA-A3 00002105): 6108 RPM CEC1: fanspeed (Fan U78CB.001.WZS00MA-A4 00002107): 6000 RPM CEC1: fanspeed (Fan U78CB.001.WZS00MA-A5 00002109): 6013 RPM CEC1: fanspeed (Fan U78CB.001.WZS00MA-A6 0000210B): 6013 RPM CEC1: fanspeed (Fan U78CB.001.WZS00MA-E1 0000210C): 4992 RPM CEC1: fanspeed (Fan U78CB.001.WZS00MA-E2 0000210D): 5016 RPM
3
Query the historical records for the CPUspeed attribute. (Power8 CEC)
renergy CEC1 CPUspeedhistory
The output of the query operation:
CEC1: CPUspeedhistory: 2027 MHZ: 20141226042900 CEC1: CPUspeedhistory: 2027 MHZ: 20141226042930 CEC1: CPUspeedhistory: 2244 MHZ: 20141226043000 CEC1: CPUspeedhistory: 2393 MHZ: 20141226043030 CEC1: CPUspeedhistory: 2393 MHZ: 20141226043100 CEC1: CPUspeedhistory: 2393 MHZ: 20141226043130 CEC1: CPUspeedhistory: 2393 MHZ: 20141226043200 CEC1: CPUspeedhistory: 2393 MHZ: 20141226043230 CEC1: CPUspeedhistory: 2393 MHZ: 20141226043300 CEC1: CPUspeedhistory: 2393 MHZ: 20141226043330 ...
4
Query all the attirbutes for management module node MM1. (For chassis)
renergy MM1 all
The output of the query operation:
mm1: availableDC: 5880W mm1: frontpaneltmp: 18.00 Centigrade mm1: inusedAC: 2848W mm1: mmtmp: 28.00 Centigrade mm1: pd1avaiablepower: 2940W mm1: pd1inusedpower: 848W mm1: pd1policy: redundantWithoutPerformanceImpact mm1: pd1powermodule1: Bay 1: 2940W mm1: pd1powermodule2: Bay 2: 2940W mm1: pd1remainpower: 1269W mm1: pd1reservedpower: 1671W mm1: pd1status: 1 - Power domain status is good. mm1: pd2avaiablepower: 2940W mm1: pd2inusedpower: 1490W mm1: pd2policy: redundantWithoutPerformanceImpact mm1: pd2powermodule1: Bay 3: 2940W mm1: pd2powermodule2: Bay 4: 2940W mm1: pd2remainpower: 51W mm1: pd2reservedpower: 2889W mm1: pd2status: 2 - Warning: Power redundancy does not exist in this power domain. mm1: thermaloutput: 9717.376000 BTU/hour
5
Query all the attirbutes for blade server node blade1.
renergy blade1 all
The output of the query operation:
blade1: CPUspeed: 4204MHZ blade1: averageDC: 227W blade1: capability: dynamicPowerMeasurement2 blade1: cappingvalue: 315W blade1: dsavingstatus: off blade1: maxCPUspeed: 4204MHZ blade1: savingstatus: off
6
Query the attributes savingstatus, cappingstatus and CPUspeed for server CEC1.
renergy CEC1 savingstatus cappingstatus CPUspeed
The output of the query operation:
CEC1: savingstatus: off CEC1: cappingstatus: on CEC1: CPUspeed: 3621 MHz
7
Turn on the power saving function of CEC1.
renergy CEC1 savingstatus=on
The output of the setting operation:
CEC1: Set savingstatus succeeded. CEC1: This setting may need some minutes to take effect.
8
Set the power capping value base on the percentage of the max-min capping value. Here, set it to 50%.
renergy CEC1 cappingperc=50
If the maximum capping value of the CEC1 is 850w, and the minimum capping value of the CEC1 is 782w, the Power Capping value will be set as ((850-782)*50% + 782) = 816w.
The output of the setting operation:
CEC1: Set cappingperc succeeded. CEC1: cappingvalue: 816
1
For more information on ‘Power System Energy Management’:
http://www-03.ibm.com/systems/power/software/energy/index.html
2
EnergyScale white paper for Power6:
http://www-03.ibm.com/systems/power/hardware/whitepapers/energyscale.html
3
EnergyScale white paper for Power7:
http://www-03.ibm.com/systems/power/hardware/whitepapers/energyscale7.html
/opt/xcat/bin/renergy
replaycons.1¶
replaycons - replay the console output for a node
The replaycons command reads the console log stored by conserver for this node, and displays it in a way that simulates the original output of the console. Using the bps value, it will throttle the speed of the output play back. (The conserver logs are stored in /var/log/consoles.)
For now, replaycons must be run locally on the system on which the console log is stored. This is normally that management node, but in a hierarchical cluster will usually be the service node.
bps]
The display rate to use to play back the console output. Default is 19200.
tail_amount
The place in the console log file to start play back, specified as the # of lines from the end.
-v|--version
Command Version.
-h|--help
Display usage message.
To replay the console for node1 at the default rate, starting 2000 lines from the end:
replaycons 19200 2000
/opt/xcat/bin/replaycons
rcons(1)|rcons.1
restartxcatd.1¶
restartxcatd - Restart the xCAT daemon (xcatd).
restartxcatd [[-h | --help] | [-v | --version] | [-r | --reload]] [-V | --verbose]
The restartxcatd command restarts the xCAT daemon (xcatd).
Linux Specific
- It will perform the xcatd fast restart. The xcatd fast restart is a specific restart which has two advantages compares to the stop and then start.
- The interval of xcatd out of service is very short.
- The in processing request which initiated by old xcatd will not be stopped by force. The old xcatd will hand over the sockets to new xcatd, but old xcat will still be waiting for the in processing request to finish before the exit.
It does the same thing as ‘service xcatd restart’ on NON-systemd enabled Operating System like rh6.x and sles11.x. But for the systemd enabled Operating System like rh7 and sles12, the ‘service xcatd restart’ just do the stop and start instead of xcatd fast restart.
It’s recommended to use restartxcatd command to restart xcatd on systemd enable system like rh7 and sles12 instead of ‘service xcatd restart’ or ‘systemctl restart xcatd’.
AIX Specific
It runs ‘stopsrc -s xcatd’ to stop xcatd first if xcatd is active, then runs ‘startsrc -s xcatd’ to start xcatd.
If the xcatd subsystem was not created, restartxcatd will create it automatically.
-h Display usage message.
-v Command Version.
-r On a Service Node, services will not be restarted.
-V Display the verbose messages.
/opt/xcat/sbin/restartxcatd
restorexCATdb.1¶
restorexCATdb - restores the xCAT db tables .
restorexCATdb [-a] [-V] [{-p | --path} path]
restorexCATdb [-b] [-V] [{-t | --timestamp} timestamp] [{-p | --path} path]
restorexCATdb [-h | --help] [-v | --version]
If not using binary restore(-b), the restorexCATdb command restores the xCAT database tables from the *.csv files in directory given by the -p flag. The site table skiptables attribute can be set to a list of tables not to restore. It will not restore isnm_perf* tables. See man dumpxCATdb.
If using the binary restore option for DB2 or postgreSQL, the entire database is restored from the binary backup made with dumpxCATdb. The database will be restored using the database Utilities. For DB2, the timestamp of the correct DB2 backup file (-t) must be provided. All applications accessing the DB2 database must be stopped before you can use the binary restore options. See the xCAT DB2 document for more information. For postgreSQL, you do not have to stop the applications accessing the database and the complete path to the backup file, must be supplied on the -p flag.
-h Display usage message.
-v Command Version.
-V Verbose.
-a All,without this flag the eventlog and auditlog will be skipped. These tables are skipped by default because restoring will generate new indexes
-b Restore from the binary image.
-p Path to the directory containing the database restore files. If restoring from the binary image (-b) and using postgeSQL, then this is the complete path to the restore file that was created with dumpxCATdb -b.
-t Use with the -b flag to designate the timestamp of the binary image to use to restore for DB2.
- To restore the xCAT database from the /dbbackup/db directory, enter:
restorexCATdb -p /dbbackup/db
- To restore the xCAT database including auditlog and eventlog from the /dbbackup/db directory, enter:
restorexCATdb -a -p /dbbackup/db
- To restore the xCAT DB2 database from the binary image with timestamp 20111130130239 enter:
restorexCATdb -b -t 20111130130239 -p /dbbackup/db
- To restore the xCAT postgreSQL database from the binary image file pgbackup.20553 created by dumpxCATdb enter:
restorexCATdb -b -p /dbbackup/db/pgbackup.20553
/opt/xcat/sbin/restorexCATdb
dumpxCATdb(1)|dumpxCATdb.1
reventlog.1¶
reventlog - retrieve or clear remote hardware event logs
reventlog noderange {number-of-entries [-s]|all [-s] | clear}
reventlog [-h | --help | -v | --version]
reventlog can display any number of remote hardware event log entries or clear them for a range of nodes. Hardware event logs are stored on each servers service processor.
number-of-entries
Retrieve the specified number of entries from the nodes’ service processors.
all
Retrieve all entries.
-s
To sort the entries from latest (always the last entry in event DB) to oldest (always the first entry in event DB). If number-of-entries specified, the latest number-of-entries events will be output in the order of latest to oldest.
clear
Clear event logs.
-h | --help
Print help.
-v | --version
Print version.
reventlog node4,node5 5
node4: SERVPROC I 09/06/00 15:23:33 Remote Login Successful User ID = USERID[00]
node4: SERVPROC I 09/06/00 15:23:32 System spn1 started a RS485 connection with us[00]
node4: SERVPROC I 09/06/00 15:22:35 RS485 connection to system spn1 has ended[00]
node4: SERVPROC I 09/06/00 15:22:32 Remote Login Successful User ID = USERID[00]
node4: SERVPROC I 09/06/00 15:22:31 System spn1 started a RS485 connection with us[00]
node5: SERVPROC I 09/06/00 15:22:32 Remote Login Successful User ID = USERID[00]
node5: SERVPROC I 09/06/00 15:22:31 System spn1 started a RS485 connection with us[00]
node5: SERVPROC I 09/06/00 15:21:34 RS485 connection to system spn1 has ended[00]
node5: SERVPROC I 09/06/00 15:21:30 Remote Login Successful User ID = USERID[00]
node5: SERVPROC I 09/06/00 15:21:29 System spn1 started a RS485 connection with us[00]
reventlog node4,node5 clear
node4: clear
node5: clear
rpower(1)|rpower.1, monstart(1)|monstart.1
rflash.1¶
rflash - Performs Licensed Internal Code (LIC) update support for HMC-attached POWER5 and POWER6 Systems, and POWER7 systems using Direct FSP management. rflash is also able to update firmware for NextScale Fan Power Controllers (FPC).
rflash [-h | --help | -v | --version]
rflash noderange -p directory {--activate concurrent | disruptive} [-V | --verbose]
rflash noderange {--commit | --recover} [-V | --verbose]
rflash noderange -p directory --activate disruptive | deferred [-d data_directory]
rflash noderange {--commit | --recover}
rflash noderange http directory
rflash The rflash command initiates Firmware updates on supported xCAT nodes. Licensed Internal Code (also known as microcode) updates are performed on supported HMC-attached POWER5 and POWER6 pSeries nodes, and POWER7 systems using Direct FSP management.
The command scans the specified directory structure for Firmware update package files applicable to the given nodes and components. And then it will automatically select the latest version for the upgrade. The firmware update files include the Microcode update package and associated XML file. They can be downloaded from the IBM Web site: http://www-933.ibm.com/support/fixcentral/.
The POWER5 and POWER6 systems contain several components that use Licensed Internal Code. The rflash command supports two of these components: the managed system (also known as the Central Electronics Complex, or CEC) and the power subsystem (also known as the Bulk Power Assembly (BPA) or Bulk Power Controller (BPC)). Some POWER5 managed systems can be attached to a power subsystem. These power subsystems can support multiple managed systems. When the rflash command is invoked, xCAT will determine the managed system or power subsystem associated with that CEC and perform the update.
The noderange can be an CEC or CEC list, a Lpar or Lpar list and a Frame or Frame list. But CEC (or Lpar) and Frame can’t be used at the same time. When the noderange is an CEC or CEC list, rflash will upgrade the firmware of the CEC or CECs in the cec list. If noderange is a Lpar or Lpar list, rflash will update Licensed Internal Code (LIC) on HMC-attached POWER5 and POWER6 pSeries nodes, and POWER7 systems using Direct FSP management. If noderange is a Frame or Frame list, rflash will update Licensed Internal Code (LIC) of the power subsystem on HMC-attached POWER5 and POWER6 pSeries nodes. The noderange can also be the specified node groups. You can specify a comma or space-separated list of node group ranges. See the noderange man page for detailed usage information.
The command will update firmware for NeXtScale FPC when given an FPC node and the http information needed to access the firmware.
The rflash command uses the xdsh command to connect to the HMC controlling the given managed system and perform the updates. Before run rflash, please use rspconfig to check if the related HMC ssh is enabled. If enable a HMC ssh connection, please use rspconfig comamnd.
Warning! This command may take considerable time to complete, depending on the number of systems being updated and the workload on the target HMC. In particular, power subsystem updates may take an hour or more if there are many attached managed systems.
Depending on the Licensed Internal Code update that is installed, the affected HMC-attached POWER5 and POWER6 systems may need to be recycled. The --activate flag determines how the affected systems activate the new code. The concurrent option activates code updates that do not require a system recycle (known as a “concurrent update”). If this option is given with an update that requires a system recycle (known as a “disruptive update”), a message will be returned, and no activation will be performed. The disruptive option will cause any affected systems that are powered on to be powered down before installing and activating the update. Once the update is complete, the command will attempt to power on any affected systems that it powered down. Those systems that were powered down when the command was issued will remain powered down when the update is complete.
The flash chip of a POWER5 and POWER6 managed system or power subsystem stores firmware in two locations, referred to as the temporary side and the permanent side. By default, most POWER5 and POWER6 systems boot from the temporary side of the flash. When the rflash command updates code, the current contents of the temporary side are written to the permanent side, and the new code is written to the temporary side. The new code is then activated. Therefore, the two sides of the flash will contain different levels of code when the update has completed.
The --commit flag is used to write the contents of the temporary side of the flash to the permanent side. This flag should be used after updating code and verifying correct system operation. The --recover flag is used to write the permanent side of the flash chip back to the temporary side. This flag should be used to recover from a corrupt flash operation, so that the previously running code can be restored.
NOTE:When the --commit or --recover two flags is used, the noderange cannot be BPA. It only can be CEC or LPAR ,and will take effect for both managed systems and power subsystems.
xCAT recommends that you shutdown your Operating System images and power off your managed systems before applying disruptive updates to managed systems or power subsystems.
Any previously activated code on the affected systems will be automatically accepted into permanent flash by this procedure.
IMPORTANT! If the power subsystem is recycled, all of its attached managed systems will be recycled.
If it outputs “Timeout waiting for prompt” during the upgrade, please set the “ppctimeout” larger in the site table. After the upgrade, remeber to change it back. If run the “rflash” command on an AIX management node, need to make sure the value of “useSSHonAIX” is “yes” in the site table.
In currently Direct FSP/BPA Management, our rflash doesn’t support concurrent value of --activate flag, and supports disruptive and deferred. The disruptive option will cause any affected systems that are powered on to be powered down before installing and activating the update. So we require that the systems should be powered off before do the firmware update.
- The deferred option will load the new firmware into the T (temp) side, but will not activate it like the disruptive firmware. The customer will continue to run the Frames and CECs working with the P (perm) side and can wait for a maintenance window where they can activate and boot the Frame/CECs with new firmware levels. Refer to the doc to get more details:
- XCAT_Power_775_Hardware_Management
In Direct FSP/BPA Management, there is -d <data_directory> option. The default value is /tmp. When do firmware update, rflash will put some related data from rpm packages in <data_directory> directory, so the execution of rflash will require available disk space in <data_directory> for the command to properly execute:
- For one GFW rpm package and one power code rpm package , if the GFW rpm package size is gfw_rpmsize, and the Power code rpm package size is power_rpmsize, it requires that the available disk space should be more than:
- 1.5*gfw_rpmsize + 1.5*power_rpmsize
For Power 775, the rflash command takes effect on the primary and secondary FSPs or BPAs almost in parallel.
- For more details about the Firmware Update using Direct FSP/BPA Management, refer to:
- XCAT_Power_775_Hardware_Management#Updating_the_BPA_and_FSP_firmware_using_xCAT_DFM
- The command will update firmware for NeXtScale FPC when given an FPC node and the http information needed to access the firmware. The http imformation required includes both the MN IP address as well as the directory containing the firmware. It is recommended that the firmware be downloaded and placed in the /install directory structure as the xCAT MN /install directory is configured with the correct permissions for http. Refer to the doc to get more details:
- XCAT_NeXtScale_Clusters
-h|--help
Writes the command’s usage statement to standard output.
-p directory
Specifies the directory where the packages are located.
-d data_directory
Specifies the directory where the raw data from rpm packages for each CEC/Frame are located. The default directory is /tmp. The option is only used in Direct FSP/BPA Management.
--activate concurrent | disruptive
Must be specified to activate the new Licensed Internal Code. The “disruptive” option will cause the target systems to be recycled. Without this flag, LIC updates will be installed only, not activated.
--commit
Used to commit the flash image in the temporary side of the chip to the permanent side for both managed systems and power subsystems.
--recover
Used to recover the flash image in the permanent side of the chip to the temporary side for both managed systems and power subsystems.
-v|--version
Displays the command’s version.
-V|--verbose
Verbose output.
1
To update only the power subsystem attached to a single HMC-attached pSeries CEC(cec_name), and recycle the power subsystem and all attached managed systems when the update is complete, and the Microcode update package and associated XML file are in /tmp/fw, enter:
rflash cec_name -p /tmp/fw --activate disruptive
2
To update only the power subsystem attached to a single HMC-attached pSeries node, and recycle the power subsystem and all attached managed systems when the update is complete, and the Microcode update package and associated XML file are in /tmp/fw, enter:
rflash bpa_name -p /tmp/fw --activate disruptive
3
To commit a firmware update to permanent flash for both managed system and the related power subsystems, enter:
rflash cec_name --commit
4
To update the firmware on a NeXtScale FPC specify the FPC node name and the HTTP location of the file including the xCAT MN IP address and the directory on the xCAT MN containing the firmware as follows:
rflash fpc01 http://10.1.147.169/install/firmware/fhet17a/ibm_fw_fpc_fhet17a-2.02_anyos_noarch.rom
/opt/xcat/bin/rflash
This command is part of the xCAT software product.
rinv(1)|rinv.1, rspconfig(1)|rspconfig.1
rinv.1¶
rinv - Remote hardware inventory
rinv [-h | --help | -v | --version]
rinv noderange {pci | model | serial | asset | vpd | mprom | deviceid | guid | firm | diag | bios | mparom | mac | all}
rinv noderange {bus | config | serial | model | firm | all}
rinv noderange {mtm | serial | mac | bios | diag | mprom | mparom | firm | all}
rinv noderange [-t]
rinv noderange [config | all]
rinv noderange [--diskpoolspace]
rinv noderange [--diskpool pool space]
rinv noderange [--fcpdevices state details]
rinv noderange [--diskpoolnames]
rinv noderange [--networknames]
rinv noderange [--network name]
rinv noderange [--ssi]
rinv noderange [--smapilevel]
rinv noderange [--wwpns fcp_channel]
rinv noderange [--zfcppool pool space]
rinv noderange [--zfcppoolnames]
rinv retrieves hardware configuration information from the on-board Service Processor for a single or range of nodes and groups.
Calling rinv for VMware will display the UUID/GUID, nuumber of CPUs, amount of memory, the MAC address and a list of Hard disks. The output for each Hard disk includes the label, size and backing file location.
pci
Retrieves PCI bus information.
bus
List all buses for each I/O slot.
config
Retrieves number of processors, speed, total memory, and DIMM locations.
model
Retrieves model number.
serial
Retrieves serial number.
firm
Retrieves firmware versions.
deconfig
Retrieves deconfigured resources. Deconfigured resources are hw components (cpus, memory, etc.) that have failed so the firmware has automatically turned those components off. This option is only capable of listing some of the deconfigured resources and should not be the only method used to check the hardware status.
-x
To output the raw information of deconfigured resources for CEC.
asset
Retrieves asset tag. Usually it’s the MAC address of eth0.
vpd
Same as specifying model, serial, deviceid, and mprom.
diag
Diagnostics information of firmware.
mprom
Retrieves mprom firmware level
deviceid
Retrieves device identification. Usually device, manufacturing and product ids.
guid
Retrieves the global unique identifier
all
All of the above.
-h | --help
Print help.
-v | --version
Print version.
-t
Set the values in the vm table to what vCenter has for the indicated nodes.
zVM specific :
--diskpoolspace
Calculates the total size of every known storage pool.--diskpool pool space
Lists the storage devices (ECKD and FBA) contained in a disk pool. Space can be: all, free, or used.--fcpdevices state details
Lists the FCP device channels that are active, free, or offline. State can be: active, free, or offline.--diskpoolnames
Lists the known disk pool names.--networknames
Lists the known network names.--network name
Shows the configuration of a given network device.--ssi
Obtain the SSI and system status.--smapilevel
Obtain the SMAPI level installed on the z/VM system.--wwpns fcp_channel
Query a given FCP device channel on a z/VM system and return a list of WWPNs.--zfcppool pool space
List the SCSI/FCP devices contained in a zFCP pool. Space can be: free or used.--zfcppoolnames
List the known zFCP pool names.
*
To retrieve all information available from blade node4, enter:
rinv node5 all node5: Machine Type/Model 865431Z node5: Serial Number 23C5030 node5: Asset Tag 00:06:29:1F:01:1A node5: PCI Information node5: Bus VendID DevID RevID Description Slot Pass/Fail node5: 0 1166 0009 06 Host Bridge 0 PASS node5: 0 1166 0009 06 Host Bridge 0 PASS node5: 0 5333 8A22 04 VGA Compatible Controller0 PASS node5: 0 8086 1229 08 Ethernet Controller 0 PASS node5: 0 8086 1229 08 Ethernet Controller 0 PASS node5: 0 1166 0200 50 ISA Bridge 0 PASS node5: 0 1166 0211 00 IDE Controller 0 PASS node5: 0 1166 0220 04 Universal Serial Bus 0 PASS node5: 1 9005 008F 02 SCSI Bus Controller 0 PASS node5: 1 14C1 8043 03 Unknown Device Type 2 PASS node5: Machine Configuration Info node5: Number of Processors: node5: Processor Speed: 866 MHz node5: Total Memory: 512 MB node5: Memory DIMM locations: Slot(s) 3 4
*
To output the raw information of deconfigured resources for CEC cec01, enter:
rinv cec01 deconfig -x cec01: <SYSTEM> <System_type>IH</System_type> <NODE> <Location_code>U78A9.001.0123456-P1</Location_code> <RID>800</RID> </NODE> </SYSTEM>
*
To retrieve ‘config’ information from the HMC-managed LPAR node3, enter:
rinv node3 config node5: Machine Configuration Info node5: Number of Processors: 1 node5: Total Memory (MB): 1024
*
To retrieve information about a VMware node vm1, enter:
rinv vm1 vm1: UUID/GUID: 42198f65-d579-fb26-8de7-3ae49e1790a7 vm1: CPUs: 1 vm1: Memory: 1536 MB vm1: Network adapter 1: 36:1b:c2:6e:04:02 vm1: Hard disk 1 (d0): 9000 MB @ [nfs_192.168.68.21_vol_rc1storage_vmware] vm1_3/vm1.vmdk vm1: Hard disk 2 (d4): 64000 MB @ [nfs_192.168.68.21_vol_rc1storage_vmware] vm1_3/vm1_5.vmdkzVM specific :
*
To list the defined network names available for a given node:
rinv pokdev61 --getnetworknamesOutput is similar to:
pokdev61: LAN:QDIO SYSTEM GLAN1 pokdev61: LAN:HIPERS SYSTEM GLAN2 pokdev61: LAN:QDIO SYSTEM GLAN3 pokdev61: VSWITCH SYSTEM VLANTST1 pokdev61: VSWITCH SYSTEM VLANTST2 pokdev61: VSWITCH SYSTEM VSW1 pokdev61: VSWITCH SYSTEM VSW2 pokdev61: VSWITCH SYSTEM VSW3
*
To list the configuration for a given network:
rinv pokdev61 --getnetwork GLAN1Output is similar to:
pokdev61: LAN SYSTEM GLAN1 Type: QDIO Connected: 1 Maxconn: INFINITE pokdev61: PERSISTENT UNRESTRICTED IP Accounting: OFF pokdev61: IPTimeout: 5 MAC Protection: Unspecified pokdev61: Isolation Status: OFF
*
To list the disk pool names available:
rinv pokdev61 --diskpoolnamesOutput is similar to:
pokdev61: POOL1 pokdev61: POOL2 pokdev61: POOL3
*
List the configuration for a given disk pool:
rinv pokdev61 --diskpool POOL1 freeOutput is similar to:
pokdev61: #VolID DevType StartAddr Size pokdev61: EMC2C4 3390-09 0001 10016 pokdev61: EMC2C5 3390-09 0001 10016
*
List the known zFCP pool names.
rinv pokdev61 --zfcppoolnamesOutput is similar to:
pokdev61: zfcp1 pokdev61: zfcp2 pokdev61: zfcp3
*
List the SCSI/FCP devices contained in a given zFCP pool:
rinv pokdev61 --zfcppool zfcp1Output is similar to:
pokdev61: #status,wwpn,lun,size,range,owner,channel,tag pokdev61: used,500512345678c411,4014412100000000,2g,3B40-3B7F,ihost13,3b77, pokdev61: used,500512345678c411,4014412200000000,8192M,3B40-3B7F,ihost13,3b77,replace_root_device pokdev61: free,500512345678c411,4014412300000000,8g,3B40-3B7F,,, pokdev61: free,5005123456789411,4014412400000000,2g,3B40-3B7F,,, pokdev61: free,5005123456789411;5005123456789411,4014412600000000,2G,3B40-3B7F,,,
rpower(1)|rpower.1
rmdef.1¶
rmdef - Use this command to remove xCAT data object definitions.
rmdef [-h | --help] [-t object-types]
rmdef [-V | --verbose] [-a | --all] [-t object-types] [-o object-names] [-f | --force] [noderange]
This command is used to remove xCAT object definitions that are stored in the xCAT database.
-a|--all
Clear the whole xCAT database. A backup of the xCAT definitions should be saved before using this option. Once all the data is removed the xCAT daemon will no longer work. Most xCAT commands will fail. In order to use xCAT commands again, you have two options. You can restore your database from your backup by switching to bypass mode, and running the restorexCATdb command. You switch to bypass mode by setting the XCATBYPASS environmant variable. (ex. “export XCATBYPASS=yes”) A second option is to run xcatconfig -d. This will restore the initial setup of the database as when xCAT was initially installed. You can then restart xcatd and run xCAT commands.
-f|--force
Use this with the all option as an extra indicator that ALL definitions are to be removed.
-h|--help
Display a usage message.
noderange
A set of comma delimited node names and/or group names. See the “noderange” man page for details on supported formats.
-o object-names
A set of comma delimited object names.
-t object-types
A set of comma delimited object types.
-V|--verbose
Verbose mode.
- To remove a range of node definitions.
rmdef -t node node1-node4
- To remove all node definitions for the nodes contained in the group bpcnodes.
rmdef -t node -o bpcnodes
- To remove the group called bpcnodes.
rmdef -t group -o bpcnodes(This will also update the values of the “groups” attribute of the member nodes.)
$XCATROOT/bin/rmdef
(The XCATROOT environment variable is set when xCAT is installed. The default value is “/opt/xcat”.)
This command is part of the xCAT software product.
mkdef(1)|mkdef.1, lsdef(1)|lsdef.1, chdef(1)|chdef.1, xcatstanzafile(5)|xcatstanzafile.5
rmdsklsnode.1¶
rmdsklsnode - Use this xCAT command to remove AIX/NIM diskless machine definitions.
rmdsklsnode [-h | --help ]
rmdsklsnode [-V|--verbose] [-f|--force] [-r|--remdef] [-i image_name] [-p|--primarySN] [-b|--backupSN] noderange
Use this command to remove all NIM client machine definitions that were created for the specified xCAT nodes.
The xCAT node definitions will not be removed. Use the xCAT rmdef command to remove xCAT node definitions.
If you are using xCAT service nodes the rmdsklsnode command will automatically determine the correct server(s) for the node and remove the NIM definitions on that server(s).
If the node you are trying to remove is currently running the rmdsklsnode command will not remove the definitions. You can use the “-f” option to shut down the node and remove the definition.
Removing alternate NIM client definitions
If you used the “-n” option when you created the NIM client definitions with the mkdsklsnode command then the NIM client machine names would be a combination of the xCAT node name and the osimage name used to initialize the NIM machine. To remove these definitions you must provide the name of the osimage that was used using the “-i” option.
In most cases you would most likely want to remove the old client definitions without disturbing the nodes that you just booted with the new alternate client definition. The rmdsklsnode -r option can be used to remove the old alternate client defintions without stopping the running node.
However, if you have NIM dump resources assign to your nodes be aware that when the old NIM alternate client definitions are removed it will leave the nodes unable to produce a system dump. This is a current limitation in the NIM support for alternate client definitions. For this reason it is recommended that you wait to do this cleanup until right before you do your next upgrade.
-f |--force
Use the force option to stop and remove running nodes. This handles the situation where a NIM machine definition indicates that a node is still running even though it is not.
-b |--backupSN
When using backup service nodes only update the backup. The default is to updat e both the primary and backup service nodes.
-h |--help
Display usage message.
-i image_name
The name of an xCAT image definition.
noderange
A set of comma delimited node names and/or group names. See the “noderange” man page for details on additional supported formats.
-p|--primarySN
When using backup service nodes only update the primary. The default is to upda te both the primary and backup service nodes.
-r|--remdef
Use this option to reset, deallocate, and remove NIM client definitions. This option will not attempt to shut down running nodes. This option should be used when remove alternate NIM client definitions that were created using mkdsklsnode -n.
-V |--verbose
Verbose mode.
- Remove the NIM client definition for the xCAT node named “node01”. Give verbose output.
rmdsklsnode -V node01
2) Remove the NIM client definitions for all the xCAT nodes in the group “aixnod es”. Attempt to shut down the nodes if they are running.
rmdsklsnode -f aixnodes
- Remove the NIM client machine definition for xCAT node “node02” that was created with the mkdsklsnode -n option and the image “AIXdskls”. (i.e. NIM client machine name “node02_AIXdskls”.)
rmdsklsnode -i AIXdskls node02
This assume that node02 is not currently running.
- Remove the old alternate client definition “node27_olddskls”.
rmdsklsnode -r -i olddskls node27
Assuming the node was booted using an new alternate NIM client definition then this will leave the node running.
/opt/xcat/bin/rmdsklsnode
This command is part of the xCAT software product.
mkdsklsnode(1)|mkdsklsnode.1
rmflexnode.1¶
rmflexnode - Delete a flexible node.
Delete a flexible node which created by the mkflexnode command.
The rmflexnode command will delete the Partition which the slots in id attribute assigned to.
The action of deleting flexible node will impact the hardware status. Before deleting it, the blades in the slot range should be in power off state.
After the deleting, use the lsflexnode to check the status of the node.
The noderange only can be a blade node.
1
Delete a flexible node base on the xCAT node blade1.
The blade1 should belong to a complex, the id attribute should be set correctly and all the slots should be in power off state.
rmflexnode blade1
/opt/xcat/bin/rmflexnode
lsflexnode(1)|lsflexnode.1, mkflexnode(1)|mkflexnode.1
rmhwconn.1¶
rmhwconn - Use this command to remove connections from CEC and Frame nodes to HMC nodes.
rmhwconn [-h| --help]
rmhwconn [-v| --version]
rmhwconn [-V| --verbose] noderange
rmhwconn noderange -T tooltype
rmhwconn -s
For PPC (with HMC) specific:
This command is used to disconnect CEC and Frame nodes from HMC nodes, according to the connection information defined in ppc talbe in xCAT DB.
Note: If a CEC belongs to a frame with a BPA installed, this CEC cannot be disconnected individually. Instead, the whole frame should be disconnected.
For PPC (without HMC, using FSPAPI) specific:
It’s used to disconnection CEC and Frame nodes from hardware server.
For PPC (use HMC as SFP) specific:
It is used to disconnect Frame nodes from HMC nodes.
-h|--help
Display usage message.
-V|--verbose
Verbose output.
-T
The tooltype is used to communicate to the CEC/Frame. The value could be lpar or fnm. The tooltype value lpar is for xCAT and fnm is for CNM.
To disconnect all CEC nodes in node group cec from their HMC nodes:
rmhwconn cec
To remove the connection for Frame node frame1:
rmhwconn frame1
To disconnect all CEC nodes in node group cec from their related hardware serveri, using lpar tooltype:
rmhwconn cec -T lpar
$XCATROOT/bin/rmhwconn
(The XCATROOT environment variable is set when xCAT is installed. The default value is “/opt/xcat”.)
This command is part of the xCAT software product.
lshwconn(1)|lshwconn.1, mkhwconn(1)|mkhwconn.1
rmhypervisor.1¶
rmhypervisor - Remove the virtualization hosts.
The rmhypervisor command can be used to remove the virtualization host.
/opt/xcat/bin/rmhypervisor
rmigrate.1¶
rmigrate - Execute migration of a guest VM between hosts/hypervisors
rmigrate noderange target_host
rmigrate noderange [destination=target_host] [action=action] [force=force] [immediate=yes_no] [max_total=total] [max_quiesce=quiesce]
rmigrate requests that a guest VM be moved from the current entity hosting it to another. It requests a live migration be done, if possible.
rmigrate migrates a VM from one z/VM member to another in an SSI cluster (only in z/VM 6.2).
destination= The name of the destination z/VM system to which the specified virtual machine will be relocated.
action= It can be: (MOVE) initiate a VMRELOCATE MOVE of the VM, (TEST) determine if VM is eligible to be relocated, or (CANCEL) stop the relocation of VM.
force= It can be: (ARCHITECTURE) attempt relocation even though hardware architecture facilities or CP features are not available on destination system, (DOMAIN) attempt relocation even though VM would be moved outside of its domain, or (STORAGE) relocation should proceed even if CP determines that there are insufficient storage resources on destination system.
immediate= It can be: (YES) VMRELOCATE command will do one early pass through virtual machine storage and then go directly to the quiesce stage, or (NO) specifies immediate processing.
max_total= The maximum wait time for relocation to complete.
max_quiesce= The maximum quiesce time a VM may be stopped during a relocation attempt.
vm table - Table governing VM paramaters. See vm(5)|vm.5 for further details. This is used to determine the current host to migrate from.
rmimage.1¶
rmimage - Removes the Linux stateless or statelite image from the file system.
rmimage [-h | –help]
rmimage [-V | –verbose] -o <OS> -a <architecture> -p <profile>
rmimage [-V | –verbose] imagename [–xcatdef]
Removes the Linux stateless or statelite image from the file system. The install dir is setup by using “installdir” attribute set in the site table.
If imagename is specified, this command uses the information in the imagenameto calculate the image root directory; otherwise, this command uses the operating system name, architecture and profile name to calculate the image root directory.
The osimage definition will not be removed from the xCAT tables by default, specifying the flag –xcatdef will remove the osimage definition, or you can use rmdef -t osimage to remove the osimage definition.
The statelite image files on the diskful service nodes will not be removed, remove the image files on the service nodes manually if necessary, for example, use command “rsync -az –delete /install <sn>:/” to remove the image files on the service nodes, where the <sn> is the hostname of the service node.
imagename specifies the name of an os image definition to be used. The specification for the image is stored in the osimage table and linuximage table.
-h | --help Display usage message.
-V | --verbose Verbose mode.
-o Operating system (fedora9,rhels5.5,sles11,etc)
-p Profile (compute,service,etc)
-a Architecture (ppc64,x86_64,etc)
--xcatdef Remove the xCAT osimage definition
- To remove a fedora9 stateless image for a compute node architecture x86_64, enter:
rmimage fedora9-x86_64-netboot-compute
or
rmimage -o fedora9 -p compute -a x86_64
- To remove a rhels5.5 statelite image for a compute node architecture ppc64 and the osimage definition, enter:
rmimage rhels5.5-ppc64-statelite-compute –xcatdef
or
rmimage -o rhels5.5 -p compute -a ppc64 –xcatdef
/opt/xcat/sbin/rmimage
This command is part of the xCAT software product.
genimage(1)|genimage.1, packimage(1)|packimage.1
rmkit.1¶
rmkit - Remove Kits from xCAT
rmkit [-? | -h | --help] [-v | --version]
rmkit [-V | --verbose] [-f | --force] [-t | --test] kitlist
The rmkit command removes kits on the xCAT management node from kit names.
Note: The xCAT support for Kits is only available for Linux operating systems.
-h|--help
Display usage message.
-V|--verbose
Verbose mode.
-v|--version
Command version.
-f|--force
Remove this kit even there is any component in this kit is listed by osimage.kitcomponents. If this option is not specified, this kit will not be removed if any kit components listed in an osimage.kitcomponents
-t|--test
Test if kitcomponents in this kit are used by osimage
kitlist
A comma delimited list of kits that are to be removed from the xCAT cluster. Each entry can be a kitname or kit basename. For kit basename, rmkit command will remove all the kits that have that kit basename.
- To remove two kits from tarball files.
rmkit kit-test1,kit-test2
Output is similar to:
Kit kit-test1-1.0-Linux,kit-test2-1.0-Linux was successfully removed.
- To remove two kits from tarball files even the kit components in them are still being used by osimages.
rmkit kit-test1,kit-test2 –force
Output is similar to:
Kit kit-test1-1.0-Linux,kit-test2-1.0-Linux was successfully removed.
- To list kitcomponents in this kit used by osimage
rmkit kit-test1,kit-test2 -t
Output is similar to:
kit-test1-kitcomp-1.0-Linux is being used by osimage osimage-test Following kitcomponents are in use: kit-test1-kitcomp-1.0-Linux
lskit(1)|lskit.1, addkit(1)|addkit.1, addkitcomp(1)|addkitcomp.1, rmkitcomp(1)|rmkitcomp.1, chkkitcomp(1)|chkkitcomp.1
~
rmkitcomp.1¶
rmkitcomp - Remove Kit components from an xCAT osimage.
rmkitcomp [-? | -h | --help] [-v | --version]
rmkitcomp [-V | --verbose] [-u | --uninstall] [-f | --force] [--noscripts] -i osimage kitcompname_list
The rmkitcomp command removes kit components from an xCAT osimage. All the kit component attribute values that are contained in the osimage will be removed, and the kit comoponent meta rpm and package rpm could be uninstalled by <-u|-****-uninstall> option.
Note: The xCAT support for Kits is only available for Linux operating systems.
-u|--uninstall
All the kit component meta rpms and package rpms in otherpkglist will be uninstalled during genimage for stateless image and updatenode for stateful nodes.
-h|--help
Display usage message.
-V|--verbose
Verbose mode.
-v|--version
Command version.
-f|--force
Remove this kit component from osimage no matter it is a dependency of other kit components.
--noscripts
Do not remove kitcomponent’s postbootscripts from osimage
-i osimage
osimage name that include this kit component.
kitcompname_list
A comma-delimited list of valid full kit component names or kit component basenames that are to be removed from the osimage. If a basename is specified, all kitcomponents matching that basename will be removed from the osimage.
- To remove a kit component from osimage
rmkitcomp -i rhels6.2-ppc64-netboot-compute comp-test1-1.0-1-rhels-6.2-ppc64
Output is similar to:
kitcomponents comp-test1-1.0-1-rhels-6.2-ppc64 were removed from osimage rhels6.2-ppc64-netboot-compute successfully
- To remove a kit component even it is still used as a dependency of other kit component.
rmkitcomp -f -i rhels6.2-ppc64-netboot-compute comp-test1-1.0-1-rhels-6.2-ppc64
Output is similar to:
kitcomponents comp-test1-1.0-1-rhels-6.2-ppc64 were removed from osimage rhels6.2-ppc64-netboot-compute successfully
- To remove a kit component from osimage and also remove the kit component meta RPM and package RPM. So in next genimage for statelss image and updatenode for statefull nodes, the kit component meta RPM and package RPM will be uninstalled.
rmkitcomp -u -i rhels6.2-ppc64-netboot-compute comp-test1-1.0-1-rhels-6.2-ppc64
Output is similar to:
kitcomponents comp-test1-1.0-1-rhels-6.2-ppc64 were removed from osimage rhels6.2-ppc64-netboot-compute successfully
lskit(1)|lskit.1, addkit(1)|addkit.1, rmkit(1)|rmkit.1, addkitcomp(1)|addkitcomp.1, chkkitcomp(1)|chkkitcomp.1
rmnimimage.1¶
rmnimimage - Use this xCAT command to remove NIM resources specified in an xCAT osimage definition.
rmnimimage [-h|--help]
rmnimimage [-V|--verbose] [-f|--force] [-d|--delete] [-x|--xcatdef] [-M|--managementnode] [-s servicenoderange] osimage_name
Use this xCAT command to remove the AIX resources specified in an xCAT osimage definition.
To list the contents of the xCAT osimage definition use the xCAT lsdef command (“lsdef -t osimage -l -o <osimage_name>”). Before running the rmnimimage command you should be absolutely certain that you really want to remove the NIM resources specified in the xCAT osimage definition!
The default behavior of this command is to remove all the NIM resources, except the lpp_source, on the xCAT management node in addition to the resources that were replicated on any xCAT service nodes.
This command may also be used to clean up individual xCAT service nodes and remove the xCAT osimage definitions.
The “nim -o remove” operation is used to remove the NIM resource definitions. If you wish to completely remove all the files and directories (left behind by the NIM command) you must specify the “-d” option when you run rmnimimage. The “-d” option will also remove the lpp_source resource.
If you wish to remove the NIM resource from one or more xCAT service nodes without removing the resources from the management node you can use the “-s <servicenoderange>” option. In this case the NIM resources specified in the xCAT osimage definition will be removed from the service nodes ONLY. The NIM resources on the management node will not be removed.
If you wish to remove NIM resources on the management node only, you can specify the “-M” option.
If you wish to also remove the xCAT osimage definition you must specify the “-x” option.
This command will not remove NIM resources if they are currently being used in another xCAT osimage definition. To see which resources are common between osimages you can specify the “-V” option. You can override this check by specifying the “-f” option.
This command will not remove NIM resources if they are currently allocated. You must deallocate the resources before they can be removed. See the xcat2nim and rmdsklsnode commands for information on how to deallocate and remove NIM machine definitions for standalone and diskless nodes.
See the AIX NIM documentation for additional details on how to deallocate and remove unwanted NIM objects.
-h |--help
Display usage message.
-d|--delete
Delete any files or directories that were left after the “nim -o remove” command was run. This option will also remove the lpp_source resouce and all files contained in the lpp_source directories. When this command completes all definitions and files will be completely erased so use with caution!
-f|--force
Override the check for shared resources when removing an xCAT osimage.
-M|--managementnode
Remove NIM resources from the xCAT management node only.
-s servicenoderange
Remove the NIM resources on these xCAT service nodes only. Do not remove the NIM resources from the xCAT management node.
osimage_name
The name of the xCAT osimage definition.
-V|--verbose
Verbose mode. This option will display the underlying NIM commands that are being called.
-x|--xcatdef
Remove the xCAT osimage definition.
- Remove all NIM resources specified in the xCAT “61image” definition.
rmnimimage 61image
The “nim -o remove” operation will be used to remove the NIM resource definitions on the management node as well as any service nodes where the resource has been replicated. This NIM operation does not completely remove all files and directories associated with the NIM resources.
- Remove all the NIM resources specified by the xCAT “61rte” osimage definition. Delete ALL files and directories associated with the NIM resources. This will also remove the lpp_source resource.
rmnimimage -d 61rte
- Remove all the NIM resources specified by the xCAT “614img” osimage definition and also remove the xCAT definition.
rmnimimage -x -d 614img
Note: When this command completes all definitions and files will be completely erased, so use with caution!
- Remove the NIM resources specified in the “614dskls” osimage definition on the xcatsn1 and xcatsn2 service nodes. Delete all files or directories associated with the NIM resources.
rmnimimage -d -s xcatsn1,xcatsn2 614dskls
- Remove the NIM resources specified in the “614old” osimage definition on the xCAT management node only.
rmnimimage -M -d 614old
/opt/xcat/bin/rmnimimage
This command is part of the xCAT software product.
mknimimage(1)|mknimimage.1
rmvlan.1¶
rmvlan - It remves the vlan from the cluster.
The rmvlan command removes the given vlan ID from the cluster. It removes the vlan id from all the swithces involved, deconfigures the nodes so that vlan adaptor (tag) will be remved, cleans up /etc/hosts, DNS and database tables for the given vlan.
For added security, the root guard and bpdu guard were enabled for the ports in this vlan by mkvlan and chvlan commands. However, the guards will not be disabled by this command. To disable them, you need to use the switch command line interface. Please refer to the switch command line interface manual to see how to disable the root guard and bpdu guard for a port.
vlanid is a unique vlan number.
To remove vlan 3
rmvlan 3
- If the nodes are KVM guest then the do the following after the vlan is removed:
rpower node1,node2 off rmvm node1,node2
/opt/xcat/bin/rmvlan
mkvlan(1)|mkvlan.1, chvlan(1)|chvlan.1, lsvlan(1)|lsvlan.1
rmvm.1¶
rmvm - Removes HMC-, DFM-, IVM-, KVM-, Vmware- and zVM-managed partitions or virtual machines.
rmvm [-h| –help]
rmvm [-v| –version]
rmvm [-V| –verbose] noderange [-r] [–service]
rmvm [-p] [-f]
rmvm noderange
The rmvm command removes the partitions specified in noderange. If noderange is an CEC, all the partitions associated with that CEC will be removed. Note that removed partitions are automatically removed from the xCAT database. For IVM-managed systems, care must be taken to not remove the VIOS partition, or all the associated partitions will be removed as well.
For DFM-managed (short For Direct FSP Management mode) normal power machines, only partitions can be removed. No options is needed.
-h Display usage message.
-v Command Version.
-V Verbose output.
-r Retain the data object definitions of the nodes.
--service Remove the service partitions of the specified CECs.
-p Purge the existence of the VM from persistant storage. This will erase all storage related to the VM in addition to removing it from the active virtualization configuration.
-p|--part Remove the specified partiton on normal power machine.
-f Force remove the VM, even if the VM appears to be online. This will bring down a live VM if requested.
- To remove the HMC-managed partition lpar3, enter:
rmvm lpar3
Output is similar to:
lpar3: Success
- To remove all the HMC-managed partitions associated with CEC cec01, enter:
rmvm cec01
Output is similar to:
lpar1: Success
lpar2: Success
lpar3: Success
- To remove the HMC-managed service partitions of the specified CEC cec01 and cec02, enter:
rmvm cec01,cec02 –service
Output is similar to:
cec01: Success
cec02: Success
- To remove the HMC-managed partition lpar1, but retain its definition, enter:
rmvm lpar1 -r
Output is similar to:
lpar1: Success
- To remove a zVM virtual machine:
rmvm gpok4
Output is similar to:
gpok4: Deleting virtual server LNX4... Done
- To remove a DFM-managed partition on normal power machine:
rmvm lpar1
Output is similar to:
lpar1: Done
/opt/xcat/bin/rmvm
mkvm(1)|mkvm.1, lsvm(1)|lsvm.1, chvm(1)|chvm.1
rmzone.1¶
rmzone - Removes a zone from the cluster.
The rmzone command is designed to remove a previously defined zone from the cluster. It will remove the zone entry in the zone table. It will remove the zone from the zonename attributes on the nodes that were assigned to the zone. Optionally, it will remove the zonename group from the nodes that were assigned to the zone. It will also remove the root ssh keys that were created for that zone on the Management Node. The rmzone command is only supported on Linux ( No AIX support). The nodes are not automatically updated with new root ssh keys by rmzone. You must run updatenode -k or xdsh -K to the nodes to update the root ssh keys. The nodes new ssh key will be assigned from the defaultzone in the zone table, or if no entries in the zone table, the keys will come from /root/.ssh. Note: if any zones in the zone table, there must be one and only one defaultzone. Otherwise, errors will occur.
-h | --help
Displays usage information.
-v | --version
Displays command version and build date.
-f | --force
Used to remove a zone that is defined as current default zone. This should only be done if you are removing all zones, or you will adding a new zone or changing an existing zone to be the default zone.
-g | --assigngroup
Remove the assigned group named zonename from all nodes assigned to the zone being removed.
-V | --Verbose
Verbose mode.
*
To remove zone1 from the zone table and the zonename attribute on all it’s assigned nodes , enter:
rmzone zone1
*
To remove zone2 from the zone table, the zone2 zonename attribute, and the zone2 group assigned to all nodes that were in zone2, enter:
rmzone zone2 -g
*
To remove zone3 from the zone table, all the node zone attributes and override the fact it is the defaultzone, enter:
rmzone zone3 -g -f
Files
/opt/xcat/bin/rmzone/
Location of the rmzone command.
L <mkzone(1)|mkzone.1>,L <chzone(1)|chzone.1>,L <xdsh(1)|xdsh.1>, updatenode(1)|updatenode.1
rnetboot.1¶
rnetboot - Cause the range of nodes to boot to network.
rnetboot [-V | --verbose] [-s boot_device_order] [-F] [-f] noderange [-m table.column==expectedstatus [-m table.col-umn=~expectedstatus]] [-t timeout] [-r retrycount]
rnetboot [-h | --help] [-v | --version]
rnetboot noderange [ipl= address]
The rnetboot command will do what is necessary to make each type of node in the given noderange boot from the network. This is usually used to boot the nodes stateless or to network install system p nodes.
-s
Set the boot device order. Accepted boot devices are hd and net.
-F
Force reboot the system no matter what state the node is. By default, rnetboot will not reboot the node if node is in ‘boot’ state.
-f
Force immediate shutdown of the partition.
-m
Use one or multiple -m flags to specify the node attributes and the expected status for the node installation monitoring and automatic retry mechanism. The operators ==, !=, =~ and !~ are valid. This flag must be used with -t flag.
Note: if the “val” fields includes spaces or any other characters that will be parsed by shell, the “attr<oper-ator>val” needs to be quoted. If the operator is ”!~”, the “attr<operator>val” needs to be quoted using single quote.
-r
specify the number of retries that the monitoring process will perform before declare the failure. The default value is 3. Setting the retrycount to 0 means only monitoring the os installation progress and will not re-initiate the installation if the node status has not been changed to the expected value after timeout. This flag must be used with -m flag.
-t
Specify the the timeout, in minutes, to wait for the expectedstatus specified by -m flag. This is a required flag if the -m flag is specified.
-V
Verbose output.
-h
Display usage message.
-v
Command Version.
rnetboot 1,3
rnetboot 14-56,70-203
rnetboot 1,3,14-56,70-203
rnetboot all,-129-256
rnetboot all -s hd,net
rnetboot all ipl=00c
nodeset(8)|nodeset.8
rollupdate.1¶
rollupdate - performs cluster rolling update
cat stanza-file | rollupdate [-V | --verbose] [-t| --test]
rollupdate [-? | -h | --help | -v | --version]
The rollupdate command creates and submits scheduler reservation jobs that will notify xCAT to shutdown a group of nodes, run optional out-of-band commands from the xCAT management node, and reboot the nodes. Currently, only LoadLeveler is supported as a job scheduler with rollupdate.
Input to the rollupdate command is passed in as stanza data through STDIN. Information such as the sets of nodes that will be updated, the name of the job scheduler, a template for generating job command files, and other control data are required. See /opt/xcat/share/xcat/rollupdate/rollupdate.input.sample and /opt/xcat/share/xcat/rollupdate/rollupdate_all.input.sample for stanza keywords, usage, and examples.
The rollupdate command will use the input data to determine each set of nodes that will be managed together as an update group. For each update group, a job scheduler command file is created and a reservation request is submitted. When the group of nodes becomes available and the scheduler activates the reservation, the xcatd daemon on the management node will be notified to begin the update process for all the nodes in the update group. If specified, prescripts will be run, an operating system shutdown command will be sent to each node, out-of-band operations can be run on the management node, and the nodes are powered back on.
The rollupdate command assumes that, if the update is to include rebooting stateless or statelite nodes to a new operating system image, the image has been created and tested, and that all relevant xCAT commands have been run for the nodes such that the new image will be loaded when xCAT reboots the nodes.
-v|--version
Command Version.
-V|--verbose
Display additional progress and error messages. Output is also logged in /var/log/xcat/rollupdate.log.
-t|--test
Run the rollupdate command in test mode only to verify the output files that are created. No scheduler reservation requests will be submitted.
-?|-h|--help
Display usage message.
To run a cluster rolling update based on the information you have created in the file /u/admin/rolling_updates/update_all.stanza enter:
cat /u/admin/rolling_updates/update_all.stanza | rollupdate
/opt/xcat/bin/rollupdate /opt/xcat/share/xcat/rollupdate/rollupdate.input.sample /opt/xcat/share/xcat/rollupdate/ll.tmpl /opt/xcat/share/xcat/rollupdate/rollupdate_all.input.sample /opt/xcat/share/xcat/rollupdate/llall.tmpl /var/log/xcat/rollupdate.log
rpower.1¶
rpower - remote power control of nodes
rpower noderange [--nodeps] [on | onstandby | off | suspend | stat | state | reset | boot] [-m table.column==expectedstatus [-m table.column=~expectedstatus]] [-t timeout] [-r retrycount]
rpower [-h | --help | -v | --version]
rpower noderange [on | off | softoff | reset | boot | stat | state | status | wake | suspend [-w timeout] [-o] [-r]]
rpower noderange [--nodeps] {of}
rpower noderange [on | off | reset | boot | onstandby]
rpower noderange [on | off | stat | state | reset | boot | of | sms | softoff]
rpower noderange [onstandby | stat | state] [-T tooltype]
rpower noderange [on | off | resetsp]
rpower noderange [rackstandby | exit_rackstandby | stat | state | resetsp]
rpower noderange [on | off | stat | state | reset | boot | of | sms]
rpower noderange [on | onstandby | off | stat | state | sms]
rpower noderange [cycle | softoff]
rpower noderange [on | off | reset | stat | softoff]
rpower controls the power for a single or range of nodes, via the out-of-band path.
on
Turn power on.
onstandby
Turn power on to standby state
-T
The value could be lpar or fnm. The tooltype value lpar is for xCAT and fnm is for CNM. The default value is “lpar”. For cold start in the large cluster, it will save a lot of time if the admins use “rpower noderange onstandby -T fnm” to power on all the CECs from the management node through the fnm connections.
rackstandby
Places the rack in the rack standby state. It requires that all CECs and DE be powered off before it will run.
exit_rackstandby
Exit Rack standby will be the default state that a rack goes into when power is initially applied to the rack. It simply moves the BPA from Rack standby to both bpa’s in standby state.
resetsp
Reboot the service processor. If there are primary and secondary FSPs/BPAs of one cec/frame, it will reboot them almost at the sametime.
softoff
Attempt to request clean shutdown of OS (may not detect failures in completing command)
off
Turn power off.
suspend
Suspend the target nodes execution.
The suspend action could be run together with -w -o -r.
Refer to the following steps to enable the suspend function:
- Add the ‘acpid’ and ‘suspend’(the suspend package is not needed on RHEL) package to the .pkglist of your osimage so that the required package could be installed correctly to your target system.
- Add two configuration files for the base function:
/etc/pm/config.d/suspend S2RAM_OPTS="--force --vbe_save --vbe_post --vbe_mode" /etc/acpi/events/suspend_event event=button/sleep.* action=/usr/sbin/pm-suspend
- Add the hook files for your specific applications which need specific action before or after the suspend action.
Refer to the ‘pm-utils’ package for how to create the specific hook files.
wake
Wake up the target nodes which is in suspend state.
Don’t try to run wake against the ‘on’ state node, it would cause the node gets to ‘off’ state.
For some of xCAT hardware such as NeXtScale, it may need to enable S3 before using wake. The following steps can be used to enable S3. Please reference pasu(1)|pasu.1 for “pasu” usage.
[root@xcatmn home]# echo "set Power.S3Enable Enable" > power-setting [root@xcatmn home]# pasu -b power-setting node01 node01: Batch mode start. node01: [set Power.S3Enable Enable] node01: Power.S3Enable=Enable node01: node01: Beginning intermediate batch update. node01: Waiting for command completion status. node01: Command completed successfully. node01: Completed intermediate batch update. node01: Batch mode completed successfully. [root@xcatmn home]# pasu node01 show all|grep -i s3 node01: IMM.Community_HostIPAddress3.1= node01: IMM.Community_HostIPAddress3.2= node01: IMM.Community_HostIPAddress3.3= node01: IMM.DNS_IP_Address3=0.0.0.0 node01: IMM.IPv6DNS_IP_Address3=:: node01: Power.S3Enable=Enable
stat | state
Print the current power state/status.
reset
Send a hard reset.
boot
If off, then power on. If on, then hard reset. This option is recommended over cycle.
cycle
Power off, then on.
of
Boot the node to open firmware console mode.
sms
Boot the node to open firmware SMS menu mode.
-m table.column==expectedstatus -m table.column=~expectedstatus
Use one or multiple -m flags to specify the node attributes and the expected status for the node installation monitoring and automatic retry mechanism. The operators ==, !=, =~ and !~ are valid. This flag must be used with -t flag.
Note: if the “val” fields includes spaces or any other characters that will be parsed by shell, the “attr<oper-ator>val” needs to be quoted. If the operator is ”!~”, the “attr<operator>val” needs to be quoted using single quote.
--nodeps
Do not use dependency table (default is to use dependency table). Valid only with on|off|boot|reset|cycle for blade power method and on|off|reset|softoff for hmc/fsp power method.
-r retrycount
specify the number of retries that the monitoring process will perform before declare the failure. The default value is 3. Setting the retrycount to 0 means only monitoring the os installation progress and will not re-initiate the installation if the node status has not been changed to the expected value after timeout. This flag must be used with -m flag.
-t timeout
Specify the the timeout, in minutes, to wait for the expectedstatus specified by -m flag. This is a required flag if the -m flag is specified.
Power off, then on.
-w timeout
To set the timeout for the suspend action to wait for the success.
-o
To specify that the target node will be power down if suspend action failed.
-r
To specify that the target node will be reset if suspend action failed.
-h | --help
Prints out a brief usage message.
-v | --version
Display the version number.
noderange(3)|noderange.3, rcons(1)|rcons.1, rinv(1)|rinv.1, rvitals(1)|rvitals.1, rscan(1)|rscan.1
rscan.1¶
rscan - Collects node information from one or more hardware control points.
The rscan command lists hardware information for each node managed by the hardware control points specified in noderange.
For the management module of blade, if the blade server is a Flex system P node, the fsp belongs to the blade server also will be scanned.
Note: The first line of the output always contains information about the hardware control point. When using the rscan command to generate output for HMC or IVM hardware control points, it provides the FSPs and BPAs as part of the output. The only exception is the rscan -u flag which provides updates made hardware control point in the xCAT database.
-h Display usage message.
-v Command Version.
-V Verbose output.
-u Updates and then prints out node definitions in the xCAT database for CEC/BPA. It updates the existing nodes that contain the same mtms and serial number for nodes managed by the specified hardware control point. This primarily works with CEC/FSP and frame/BPA nodes when the node name is not the same as the managed system name on hardware control point (HMC), This flag will update the BPA/FSP node name definitions to be listed as the managed system name in the xCAT database.
For the Flex system manager, both the blade server and fsp object of xCAT will be updated if the mpa and slot id are matched to the object which has been defined in the xCAT database.
Note: only the matched object will be updated.
-w Writes output to xCAT database.
-x XML format.
-z Stanza formated output.
0 The command completed successfully.
1 An error has occurred.
- To list all nodes managed by HMC hmc01 in tabular format, enter:
rscan hmc01
Output is similar to:
type name id type-model serial-number address
hmc hmc01 7310-C05 10F426A hmc01
fsp Server-9117-MMA-SN10F6F3D 9117-MMA 10F6F3D 3.3.3.197
lpar lpar3 4 9117-MMA 10F6F3D
lpar lpar2 3 9117-MMA 10F6F3D
lpar lpar1 2 9117-MMA 10F6F3D
lpar p6vios 1 9117-MMA 10F6F3D
- To list all nodes managed by IVM ivm02 in XML format and write the output to the xCAT database, enter:
rscan ivm02 -x -w
Output is similar to:
<Node>
<cons></cons>
<profile></profile>
<parent></parent>
<serial>10B7D1G</serial>
<model>9133-55A</model>
<node>Server-9133-55A-10B7D1G</node>
<mgt>ivm</mgt>
<nodetype>fsp</nodetype>
<hcp>ivm02</hcp>
<groups>fsp,all</groups>
<id>10</id>
</Node>
<Node>
<cons>ivm</cons>
<profile>lpar01</profile>
<parent>Server-9133-55A-10B7D1G</parent>
<serial></serial>
<model></model>
<node>lpar01</node>
<mgt>ivm</mgt>
<nodetype>lpar,osi</nodetype>
<hcp>ivm02</hcp>
<groups>lpar,all</groups>
<id>1</id>
<Node>
</Node>
<cons>ivm</cons>
<profile>lpar02</profile>
<parent>Server-9133-55A-10B7D1G</parent>
<serial></serial>
<model></model>
<node>lpar02</node>
<mgt>ivm</mgt>
<nodetype>lpar,osi</nodetype>
<hcp>ivm02</hcp>
<groups>lpar,all</groups>
<id>2</id>
</Node>
- To list all nodes managed by HMC hmc02 in stanza format and write the output to the xCAT database, enter:
rscan hmc02 -z -w
Output is similar to:
Server-9458-100992001Y_B:
objtype=node
nodetype=bpa
id=2
model=9458-100
serial=992001Y
hcp=hmc02
profile=
parent=
groups=bpa,all
mgt=hmc
cons=
Server-9119-590-SN02C5F9E:
objtype=node
type=fsp
id=10
model=9119-590
serial=02C5F9E
hcp=hmc02
profile=
parent=Server-9458-100992001Y_B
groups=fsp,all
mgt=hmc
cons=
lpar01:
objtype=node
nodetype=lpar,osi
id=1
model=
serial=
hcp=hmc02
profile=lpar01
parent=Server-9119-590-SN02C5F9E
groups=lpar,all
mgt=hmc
cons=hmc
lpar02:
objtype=node
nodetype=lpar,osi
id=2
model=
serial=
hcp=hmc02
profile=lpar02
parent=Server-9119-590-SN02C5F9E
groups=lpar,all
mgt=hmc
cons=hmc
- To update definitions of nodes, which is managed by hmc03, enter:
rscan hmc03 -u
Output is similar to:
#Updated following nodes:
type name id type-model serial-number address
fsp Server-9125-F2A-SN0262672-B 3 9125-F2A 0262672 192.168.200.243
- To collects the node information from one or more hardware control points on zVM AND populate the database with details collected by rscan:
rscan gpok2 -W
Output is similar to:
gpok2:
objtype=node
arch=s390x
os=sles10sp3
hcp=gpok3.endicott.ibm.com
userid=LINUX2
nodetype=vm
parent=POKDEV61
groups=all
mgt=zvm
- To scan the Flex system cluster:
rscan cmm01
Output is similar to:
type name id type-model serial-number mpa address
cmm AMM680520153 0 789392X 100048A cmm01 cmm01
blade SN#YL10JH184067 1 789542X 10F752A cmm01 12.0.0.9
xblade SN#YL10JH184068 2 789542X 10F652A cmm01 12.0.0.10
blade SN#YL10JH184079 3 789542X 10F697A cmm01 12.0.0.11
- To update the Flex system cluster:
rscan cmm01 -u
Output is similar to:
cmm [AMM680520153] Matched To =>[cmm01]
blade [SN#YL10JH184067] Matched To =>[cmm01node01]
blade [SN#YL10JH184079] Matched To =>[cmm01node03]
/opt/xcat/bin/rscan
lsslp(1)|lsslp.1
rsetboot.1¶
rsetboot noderange {hd | net | cd | default | stat}
rsetboot [-h | --help | -v | --version]
rsetboot sets the boot media that should be used on the next boot of the specified nodes. After the nodes are booted with the specified device (e.g. via rpower(1)|rpower.1), the nodes will return to using the default boot device specified in the BIOS. Currently this command is only supported for IPMI nodes.
hd
Boot from the hard disk.
net
Boot over the network, using a PXE or BOOTP broadcast.
cd
Boot from the CD or DVD drive.
def | default
Boot using the default set in BIOS.
stat
Display the current boot setting.
Set nodes 1 and 3 to boot from the network on the next boot:
rsetboot node1,node3 net
Display the next-boot value for nodes 14-56 and 70-203:
rsetboot node[14-56],node[70-203] stat
Restore the next-boot value for these nodes back to their default set in the BIOS:
rsetboot node1,node3,node[14-56],node[70-203] default
rbootseq(1)|rbootseq.1
rspconfig.1¶
rspconfig - Configures nodes’ service processors
rspconfig [-h | --help | -v | --version]
rspconfig noderange {alert | snmpdest | community}
rspconfig noderange alert={on | enable | off | disable}
rspconfig noderange snmpdest=snmpmanager-IP
rspconfig noderange community={public|string}
rspconfig noderange {ip | netmask | gateway | backupgateway | garp}
rspconfig noderange garp={time}
rspconfig noderange {sshcfg | snmpcfg | pd1 | pd2 | network | swnet | ntp | textid | frame}
rspconfig noderange USERID={newpasswd} updateBMC={y | n}
rspconfig noderange sshcfg={enable | disable}
rspconfig noderange snmpcfg={enable | disable}
rspconfig noderange solcfg={enable | disable}
rspconfig noderange pd1={nonred | redwoperf | redwperf}
rspconfig noderange pd2={nonred | redwoperf | redwperf}
rspconfig noderange network={[ip],[host],[gateway],[netmask]|\*}
rspconfig noderange initnetwork={[ip],[host],[gateway],[netmask]|\*}
rspconfig noderange textid={\*|textid}
rspconfig singlenode frame={frame_number}
rspconfig noderange frame={\*}
rspconfig noderange swnet={[ip],[gateway],[netmask]}
rspconfig noderange ntp={[ntpenable],[ntpserver],[frequency],[v3]}
rspconfig noderange {autopower | iocap | decfg | memdecfg | procdecfg | time | date | spdump | sysdump | network}
rspconfig noderange autopower={enable | disable}
rspconfig noderange iocap={enable | disable}
rspconfig noderange time={hh:mm:ss}
rspconfig noderange date={mm:dd:yyyy}
rspconfig noderange decfg={enable|disable:policyname,...}
rspconfig noderange procdecfg={configure|deconfigure:processingunit:id,...}
rspconfig noderange memdecfg={configure|deconfigure:processingunit:unit|bank:id,...>}
rspconfig noderange network={nic,\*}
rspconfig noderange network={nic,[IP,][hostname,][gateway,][netmask]}
rspconfig noderange network={nic,0.0.0.0}
rspconfig noderange HMC_passwd={currentpasswd,newpasswd}
rspconfig noderange admin_passwd={currentpasswd,newpasswd}
rspconfig noderange general_passwd={currentpasswd,newpasswd}
rspconfig noderange \*_passwd={currentpasswd,newpasswd}
rspconfig noderange {hostname}
rspconfig noderange hostname={\*|name}
rspconfig noderange --resetnet
rspconfig noderange sshcfg={enable | disable}
rspconfig noderange snmpcfg={enable | disable}
rspconfig noderange network={[ip],[host],[gateway],[netmask]|\*}
rspconfig noderange solcfg={enable | disable}
rspconfig noderange textid={\*|textid}
rspconfig noderange cec_off_policy={poweroff | stayon}
rspconfig noderange {network}
rspconfig noderange network={nic,\*}
rspconfig noderange network={nic,[IP,][hostname,][gateway,][netmask]}
rspconfig noderange network={nic,0.0.0.0}
rspconfig noderange HMC_passwd={currentpasswd,newpasswd}
rspconfig noderange admin_passwd={currentpasswd,newpasswd}
rspconfig noderange general_passwd={currentpasswd,newpasswd}
rspconfig noderange \*_passwd={currentpasswd,newpasswd}
rspconfig noderange {hostname}
rspconfig noderange hostname={\*|name}
rspconfig noderange --resetnet
rspconfig noderange HMC_passwd={currentpasswd,newpasswd}
rspconfig noderange admin_passwd={currentpasswd,newpasswd}
rspconfig noderange general_passwd={currentpasswd,newpasswd}
rspconfig noderange \*_passwd={currentpasswd,newpasswd}
rspconfig noderange {sysname}
rspconfig noderange sysname={\* | name}
rspconfig noderange {pending_power_on_side}
rspconfig noderange pending_power_on_side={temp | perm}
rspconfig noderange {cec_off_policy}
rspconfig noderange cec_off_policy={poweroff | stayon}
rspconfig noderange {BSR}
rspconfig noderange {huge_page}
rspconfig noderange huge_page={NUM}
rspconfig noderange {setup_failover}
rspconfig noderange setup_failover={enable | disable}
rspconfig noderange {force_failover}
rspconfig noderange --resetnet
rspconfig noderange HMC_passwd={currentpasswd,newpasswd}
rspconfig noderange admin_passwd={currentpasswd,newpasswd}
rspconfig noderange general_passwd={currentpasswd,newpasswd}
rspconfig noderange \*_passwd={currentpasswd,newpasswd}
rspconfig noderange {frame}
rspconfig noderange frame={\*|frame_number}
rspconfig noderange {sysname}
rspconfig noderange sysname={\* | name}
rspconfig noderange {pending_power_on_side}
rspconfig noderange pending_power_on_side={temp | perm}
rspconfig noderange --resetnet
rspconfig noderange {sshcfg}
rspconfig noderange sshcfg={enable | disable}
rspconfig noderange --resetnet
rspconfig configures various settings in the nodes’ service processors. If only a keyword is specified, without the =, it displays the current value.
For options autopower | iocap | decfg | memdecfg | procdecfg | time | date | spdump | sysdump | network, user need to use chdef -t site enableASMI=yes to enable ASMI first.
Turn on or off SNMP alerts.
autopower={enable|disable}
Select the policy for auto power restart. If enabled, the system will boot automatically once power is restored after a power disturbance.
backupgateway
Get the BMC backup gateway ip address.
community={public|string}
Get or set the SNMP commmunity value. The default is public.
date={mm:dd:yyy}
Enter the current date.
decfg={enable|disable:policyname,...}
Enables or disables deconfiguration policies.
frame={framenumber|\*}
Set or get frame number. If no framenumber and * specified, framenumber for the nodes will be displayed and updated in xCAAT database. If framenumber is specified, it only supports single node and the framenumber will be set for that frame. If * is specified, it supports noderange and all the frame numbers for the noderange will be read from xCAT database and set to frames. Setting the frame number is a disruptive command which requires all CECs to be powered off prior to issuing the command.
cec_off_policy={poweroff | stayon}
Set or get cec off policy after lpars are powered off. If no cec_off_policy value specified, the cec_off_policy for the nodes will be displayed. the cec_off_policy has two values: poweroff and stayon. poweroff means Power off when last partition powers off. stayon means Stay running after last partition powers off. If cec_off_policy value is specified, the cec off policy will be set for that cec.
HMC_passwd={currentpasswd,newpasswd}
Change the password of the userid HMC for CEC/Frame. If the CEC/Frame is the factory default, the currentpasswd should NOT be specified; otherwise, the currentpasswd should be specified to the current password of the userid HMC for the CEC/Frame.
admin_passwd={currentpasswd,newpasswd}
Change the password of the userid admin for CEC/Frame from currentpasswd to newpasswd. If the CEC/Frame is the factory default, the currentpasswd should NOT be specified; otherwise, the currentpasswd should be specified to the current password of the userid admin for the CEC/Frame.
general_passwd={currentpasswd,newpasswd}
Change the password of the userid general for CEC/Frame from currentpasswd to newpasswd. If the CEC/Frame is the factory default, the currentpasswd should NOT be specified; otherwise, the currentpasswd should be specified to the current password of the userid general for the CEC/Frame.
** \*_passwd**={currentpasswd,newpasswd}
Change the passwords of the userids HMC, admin and general for CEC/Frame from currentpasswd to newpasswd. If the CEC/Frame is the factory default, the currentpasswd should NOT be specified; otherwise, if the current passwords of the userids HMC, admin and general for CEC/Frame are the same one, the currentpasswd should be specified to the current password, and then the password will be changed to the newpasswd. If the CEC/Frame is NOT the factory default, and the current passwords of the userids HMC, admin and general for CEC/Frame are NOT the same one, this option could NOT be used, and we should change the password one by one.
frequency
The NTP update frequency (in minutes).
garp=time
Get or set Gratuitous ARP generation interval. The unit is number of 1/2 second.
gateway
The gateway ip address.
hostname
Display the CEC/BPA system names.
BSR
Get Barrier Synchronization Register (BSR) allocation for a CEC.
huge_page
Query huge page information or request NUM of huge pages for CEC. If no value specified, it means query huge page information for the specified CECs, if a CEC is specified, the specified huge_page value NUM will be used as the requested number of huge pages for the CEC, if CECs are specified, it means to request the same NUM huge pages for all the specified CECs.
setup_failover={enable|disable}
Enable or disable the service processor failover function of a CEC or display status of this function.
force_failover
Force a service processor failover from the primary service processor to the secondary service processor.
hostname={\*|name}
Set CEC/BPA system names to the names in xCAT DB or the input name.
iocap={enable|disable}
Select the policy for I/O Adapter Enlarged Capacity. This option controls the size of PCI memory space allocated to each PCI slot.
ip
The ip address.
memdecfg={configure|deconfigure:processingunit:unit|bank:id,...}
Select whether each memory bank should be enabled or disabled. State changes take effect on the next platform boot.
netmask
The subnet mask.
network={[ip],[host],[gateway],[netmask]|*}
For MPA: get or set the MPA network parameters. If ‘*’ is specified, all parameters are read from the xCAT database.
For FSP of Flex system P node: set the network parameters. If ‘*’ is specified, all parameters are read from the xCAT database.
initnetwork={[ip],[host],[gateway],[netmask]|*}
For MPA only. Connecting to the IP of MPA from the hosts.otherinterfaces to set the MPA network parameters. If ‘*’ is specified, all parameters are read from the xCAT database.
network={nic,{[ip],[host],[gateway],[netmask]}|*}
Not only for FSP/BPA but also for IMM. Get or set the FSP/BPA/IMM network parameters. If ‘*’ is specified, all parameters are read from the xCAT database. If the value of ip is ‘0.0.0.0’, this nic will be configured as a DHCP client. Otherwise this nic will be configured with a static IP.
Note that IPs of FSP/BPAs will be updated with this option, user needs to put the new IPs to /etc/hosts manually or with xCAT command makehosts. For more details, see the man page of makehosts.
nonred
Allows loss of redundancy.
ntp={[ntpenable],[ntpserver],[frequency],[v3]}
Get or set the MPA Network Time Protocol (NTP) parameters.
ntpenable
Enable or disable NTP (enable|disable).
ntpserver
Get or set NTP server IP address or name.
pd1={nonred | redwoperf | redwperf}
Power Domain 1 - determines how an MPA responds to a loss of redundant power.
pd2={nonred | redwoperf | redwperf}
Power Domain 2 - determines how an MPA responds to a loss of redundant power.
procdecfg={configure|deconfigure:processingunit:id,...}
Selects whether each processor should be enabled or disabled. State changes take effect on the next platform boot.
redwoperf
Prevents components from turning on that will cause loss of power redundancy.
redwperf
Power throttles components to maintain power redundancy and prevents components from turning on that will cause loss of power redundancy.
snmpcfg={enable|disable}
Enable or disable SNMP on MPA.
snmpdest=snmpmanager-IP
Get or set where the SNMP alerts should be sent to.
solcfg={enable|disable}
Enable or disable the sol on MPA (or CMM) and blade servers belongs to it.
spdump
Performs a service processor dump.
sshcfg={enable|disable}
Enable or disable SSH on MPA.
swnet={[ip],[gateway],[netmask]}
Set the Switch network parameters.
sysdump
Performs a system dump.
sysname
Query or set sysname for CEC or Frame. If no value specified, means to query sysname of the specified nodes. If ‘*’ specified, it means to set sysname for the specified nodes, and the sysname values would get from xCAT datebase. If a string is specified, it means to use the string as sysname value to set for the specified node.
pending_power_on_side={temp|perm}
List or set pending power on side for CEC or Frame. If no pending_power_on_side value specified, the pending power on side for the CECs or frames will be displayed. If specified, the pending_power_on_side value will be set to CEC’s FSPs or Frame’s BPAs. The value ‘temp’ means T-side or temporary side. The value ‘perm’ means P-side or permanent side.
time={hh:mm:ss}
Enter the current time in UTC (Coordinated Universal Time) format.
textid={\*|textid}
Set the blade or MPA textid. When using ‘*’, the textid used is the node name specified on the command-line. Note that when specifying an actual textid, only a single node can be specified in the noderange.
USERID={newpasswd} updateBMC={y|n}
Change the password of the userid USERID for CMM in Flex system cluster. The option updateBMC can be used to specify whether updating the password of BMCs that connected to the speified CMM. The value is ‘y’ by default which means whenever updating the password of CMM, the password of BMCs will be also updated. Note that there will be several seconds needed before this command complete.
If value \* is specified for USERID and the object node is Flex System X node, the password used to access the BMC of the System X node through IPMI will be updated as the same password of the userid USERID of the CMM in the same cluster.
--resetnet
Reset the network interfaces of the specified nodes.
v3
Enable or disable v3 authentication (enable|disable).
-h | --help
Prints out a brief usage message.
-v, --version
Display the version number.
*
To setup new ssh keys on the Management Module mm:
B<rspconfig> mm snmpcfg=enable sshcfg=enable
*
To turn on SNMP alerts for node5:
rspconfig node5 alert=on
node5: Alerts: enabled
*
To display the destination setting for SNMP alerts for node4:
rspconfig node4 snmpdest
node4: BMC SNMP Destination 1: 9.114.47.227
*
To display the frame number for frame 9A00-10000001
rspconfig 9A00-10000001 frame
9A00-10000001: 1
*
To set the frame number for frame 9A00-10000001
rspconfig 9A00-10000001 frame=2
9A00-10000001: SUCCESS
*
To set the frame numbers for frame 9A00-10000001 and 9A00-10000002
rspconfig 9A00-10000001,9A00-10000002 frame=\*
9A00-10000001: SUCCESS 9A00-10000002: SUCCESS
*
To display the MPA network parameters for mm01:
rspconfig mm01 network
mm01: MM IP: 192.168.1.47 mm01: MM Hostname: MM001125C31F28 mm01: Gateway: 192.168.1.254 mm01: Subnet Mask: 255.255.255.224
*
To change the MPA network parameters with the values in the xCAT database for mm01:
rspconfig mm01 network=\*
mm01: MM IP: 192.168.1.47 mm01: MM Hostname: mm01 mm01: Gateway: 192.168.1.254 mm01: Subnet Mask: 255.255.255.224
*
To change only the gateway parameter for the MPA network mm01:
rspconfig mm01 network=,,192.168.1.1,
mm01: Gateway: 192.168.1.1
*
To display the FSP network parameters for fsp01:
rspconfig fsp01 network
fsp01: eth0: IP Type: Dynamic IP Address: 192.168.1.215 Hostname: Gateway: Netmask: 255.255.255.0 eth1: IP Type: Dynamic IP Address: 192.168.200.51 Hostname: fsp01 Gateway: Netmask: 255.255.255.0
*
To change the FSP network parameters with the values in command line for eth0 on fsp01:
rspconfig fsp01 network=eth0,192.168.1.200,fsp01,,255.255.255.0
fsp01: Success to set IP address,hostname,netmask
*
To change the FSP network parameters with the values in the xCAT database for eth0 on fsp01:
rspconfig fsp01 network=eth0,\*
fsp01: Success to set IP address,hostname,gateway,netmask
*
To configure eth0 on fsp01 to get dynamic IP address from DHCP server:
rspconfig fsp01 network=eth0,0.0.0.0
fsp01: Success to set IP type to dynamic.
*
To get the current power redundancy mode for power domain 1 on mm01:
rspconfig mm01 pd1
mm01: Redundant without performance impact
*
To change the current power redundancy mode for power domain 1 on mm01 to non-redundant:
rspconfig mm01 pd1=nonred
mm01: nonred
*
To enable NTP with an NTP server address of 192.168.1.1, an update frequency of 90 minutes, and with v3 authentication enabled on mm01:
rspconfig mm01 ntp=enable,192.168.1.1,90,enable
mm01: NTP: disabled mm01: NTP Server: 192.168.1.1 mm01: NTP: 90 (minutes) mm01: NTP: enabled
*
To disable NTP v3 authentication only on mm01:
rspconfig mm01 ntp=,,,disable
mm01: NTP v3: disabled
*
To disable Predictive Failure and L2 Failure deconfiguration policies on mm01:
rspconfig mm01 decfg=disable:predictive,L3
mm01: Success
*
To deconfigure processors 4 and 5 of Processing Unit 0 on mm01:
rspconfig mm01 procedecfg=deconfigure:0:4,5
mm01: Success
*
To check if CEC sysname set correct on mm01:
rspconfig mm01 sysname
mm01: mm01rspconfig mm01 sysname=cec01
mm01: Successrspconfig mm01 sysname
mm01: cec01
*
To check and change the pending_power_on_side value of cec01’s fsps:
rspconfig cec01 pending_power_on_side
cec01: Pending Power On Side Primary: temp cec01: Pending Power On Side Secondary: temprspconfig cec01 pending_power_on_side=perm
cec01: Successrspconfig cec01 pending_power_on_side
cec01: Pending Power On Side Primary: perm cec01: Pending Power On Side Secondary: perm
*
To show the BSR allocation for cec01:
rspconfig cec01 BSR
cec01: Barrier Synchronization Register (BSR) cec01: Number of BSR arrays: 256 cec01: Bytes per BSR array : 4096 cec01: Available BSR array : 0 cec01: Partition name: BSR arrays cec01: lpar01 : 32 cec01: lpar02 : 32 cec01: lpar03 : 32 cec01: lpar04 : 32 cec01: lpar05 : 32 cec01: lpar06 : 32 cec01: lpar07 : 32 cec01: lpar08 : 32
*
To query the huge page information for CEC1, enter:
rspconfig CEC1 huge_page
CEC1: Huge Page Memory CEC1: Available huge page memory(in pages): 0 CEC1: Configurable huge page memory(in pages): 12 CEC1: Page Size (in GB): 16 CEC1: Maximum huge page memory(in pages): 24 CEC1: Requested huge page memory(in pages): 15 CEC1: Partition name: Huge pages CEC1: lpar1 : 3 CEC1: lpar5 : 3 CEC1: lpar9 : 3 CEC1: lpar13 : 3 CEC1: lpar17 : 0 CEC1: lpar21 : 0 CEC1: lpar25 : 0 CEC1: lpar29 : 0
*
To request 10 huge pages for CEC1, enter:
rspconfig CEC1 huge_page=10
CEC1: Success
*
To disable service processor failover for cec01, in order to complete this command, the user should power off cec01 first:
rspconfig cec01 setup_failover
cec01: Failover status: Enabledrpower cec01 off
rspconfig cec01 setup_failover=disable
cec01: Successrspconfig cec01 setup_failover
cec01: Failover status: Disabled
*
To force service processor failover for cec01:
lshwconn cec01
cec01: 192.168.1.1: LINE DOWN cec01: 192.168.2.1: sp=primary,ipadd=192.168.2.1,alt_ipadd=unavailable,state=LINE UP cec01: 192.168.1.2: sp=secondary,ipadd=192.168.1.2,alt_ipadd=unavailable,state=LINE UP cec01: 192.168.2.2: LINE DOWNrspconfig cec01 force_failover
cec01: Success.lshwconn cec01
cec01: 192.168.1.1: sp=secondary,ipadd=192.168.1.1,alt_ipadd=unavailable,state=LINE UP cec01: 192.168.2.1: LINE DOWN cec01: 192.168.1.2: LINE DOWN cec01: 192.168.2.2: sp=primary,ipadd=192.168.2.2,alt_ipadd=unavailable,state=LINE UP
*
To deconfigure memory bank 9 and 10 of Processing Unit 0 on mm01:
rspconfig mm01 memdecfg=deconfigure:bank:0:9,10
mm01: Success
*
To reset the network interface of the specified nodes:
rspconfig -***-resetnet*
Output is similar to:
Start to reset network.. Reset network failed nodes: Reset network succeed nodes: Server-8233-E8B-SN1000ECP-A,Server-9119-FHA-SN0275995-B,Server-9119-FHA-SN0275995-A, Reset network finished.
*
To update the existing admin password on fsp:
rspconfig fsp admin_passwd=admin,abc123
fsp: Success
*
To set the initial password for user HMC on fsp:
rspconfig fsp HMC_passwd=,abc123
fsp: Success
noderange(3)|noderange.3, rpower(1)|rpower.1, rcons(1)|rcons.1, rinv(1)|rinv.1, rvitals(1)|rvitals.1, rscan(1)|rscan.1, rflash(1)|rflash.1
rspreset.1¶
rspreset - resets the service processors associated with the specified nodes
rspreset resets the service processors associated with the specified nodes. It searches the nodehm table and associated tables to find the service processors associated with the nodes specified. If the node is a BMC-based node, the node’s BMC will be reset. If the node is a blade, the blade’s on board service processor will be reset.
Reset the service processor that controls node5:
rspreset node5
rpower(1)|rpower.1, nodehm(5)|nodehm.5
rvitals.1¶
rvitals - remote hardware vitals
rvitals [-h | --help | -v | --version]
rvitals noderange {temp | voltage | lcds | all}
rvitals noderange {rackenv | lcds | all} [-V| --verbose]
rvitals noderange {temp | voltage | wattage | fanspeed | power | leds | summary | all}
rvitals noderange {temp | wattage | fanspeed | leds | summary | all}
rvitals noderange {temp | voltage | wattage | fanspeed | power | leds | all}
rvitals retrieves hardware vital information from the on-board Service Processor for a single or range of nodes and groups.
cputemp
Retrieves CPU temperatures.
disktemp
Retrieves HD back plane temperatures.
ambtemp
Retrieves ambient temperatures.
temp
Retrieves all temperatures.
voltage
Retrieves power supply and VRM voltage readings.
fanspeed
Retrieves fan speeds.
lcds
Retrieves LCDs status.
rackenv
Retrieves rack environmentals.
leds
Retrieves LEDs status.
power
Retrieves power status.
powertime
Retrieves total power uptime. This value only increases, unless the Service Processor flash gets updated.
reboot
Retrieves total number of reboots. This value only increases, unless the Service Processor flash gets updated.
state
Retrieves the system state.
all
All of the above.
-h | --help
Print help.
-v | --version
Print version.
rvitals node5 all
node5: CPU 1 Temperature: + 29.00 C (+ 84.2 F)
node5: CPU 2 Temperature: + 19.00 C (+ 66.2 F)
node5: DASD Sensor 1 Temperature: + 32.00 C (+ 89.6 F)
node5: System Ambient Temperature Temperature: + 26.00 C (+ 78.8 F)
node5: +5V Voltage: + 5.01V
node5: +3V Voltage: + 3.29V
node5: +12V Voltage: + 11.98V
node5: +2.5V Voltage: + 2.52V
node5: VRM1 Voltage: + 1.61V
node5: VRM2 Voltage: + 1.61V
node5: Fan 1 Percent of max: 100%
node5: Fan 2 Percent of max: 100%
node5: Fan 3 Percent of max: 100%
node5: Fan 4 Percent of max: 100%
node5: Fan 5 Percent of max: 100%
node5: Fan 6 Percent of max: 100%
node5: Current Power Status On
node5: Current LCD1: SuSE Linux
node5: Power On Seconds 11855915
node5: Number of Reboots 930
node5: System State Booting OS or in unsupported OS
rpower(1)|rpower.1, rinv(1)|rinv.1
sinv.1¶
sinv - Checks the software configuration of the nodes in the cluster.
sinv [-o output] [-p template path] [-t template count] [-s seed node] [-i] [-e] [-r] [-V] [--devicetype type_of_device] [-l userID] [[-f command file] | [-c command]]
sinv [-h | -v]
The sinv command is designed to check the configuration of the nodes in a cluster. The command takes as input command line flags, and one or more templates which will be compared against the output of the xdsh command, designated to be run by the -c or -f flag, on the nodes in the noderange.
The nodes will then be grouped according to the template they match and a report returned to the administrator in the output file designated by the -o flag, or to stdout.
sinv supports checking the output from the rinv or xdsh command.
The sinv command is an xCAT Distributed Shell Utility.
COMMAND SPECIFICATION:
- The xdsh or rinv command to execute on the remote targets is specified by the
- -c flag, or by the -f flag
which is followed by the fully qualified path to a file containing the command.
Note: do not add | xdshcoll to the command on the command line or in the command file, it is automatically added by sinv.
The syntax for the -c sinv parameter is as follows:
“command[; command]...”
where command is the command to run on the remote target. Quotation marks are required to ensure that all commands in the list are executed remotely, and that any special characters are interpreted correctly on the remote target.
The sinv command does not work with any interactive commands, including those that read from standard input.
REMOTE SHELL COMMAND:
For xdsh, support is explicitly provided for AIX Remote Shell and OpenSSH, but any secure remote command that conforms to the IETF (Internet Engineering Task Force) Secure Remote Command Protocol can be used. See man xdsh for more details.
-o | --output report output file
Optional output file. This is the location of the file that will contain the report of the nodes that match, and do not match, the input templates. If the flag is not used, the output will go to stdout.
-p | --tp template path
This is the path to the template file. The template contains the output of xdsh command, that has been run against a “seed” node, a node that contains the configuration that you would like all nodes in your noderange to match.
The admin can create the template by running the xdsh command on the seed node, pipe to xdshcoll ( required) and store the output in the template path. See examples.
Note: The admin can also edit the template to remove any lines that they do not want checked.
An alternative method is to use the [-s seed node] parameter, which will automatically build the template for you from the seed node named.
If a template path file does not exist, and a seed node is not input, then sinv will automatically use the one node in the noderange as the seed node and build the template.
-t | --tc template count
This count is the number of templates that the command will use to check for nodes matches. If the template in the template path does not match a node, the sinv will check additional templates up to the template count.
For each node, it will compare the node against each template to see if there is a match. If there is no match, and we are not over the template count, then a new template will be created from the node output. This will result in having all nodes that match a given template reported in their group at the end of the run in the output file. If no template count is specified, 0 is the default, and all nodes will be compared against the first template.
-s | --seed seed node
This is the node that will be used to build the first template that is stored in template path. You can use this parameter instead of running the command yourself to build the template.
Note: If the template path file does not exists, and no seed node is supplied, the seed node automatically is one node in the noderange.
-i | --ignorefirst
This flag suppresses the reporting of the nodes matching the first template. In very large systems, you may not want to show the nodes that have the correct configuration, since the list could contain thousands of nodes. This allows you to only report the nodes that do not match the required configuration.
-e | --exactmatch
This requires the check of node output against template to be an exact match. If this flag is not set, sinv checks to see if the return from the xdsh command to the nodes contain a match for each line in the input template (except for xdshcoll header and comments). If not in exactmatch mode, there can exist more lines in the xdsh return from the nodes.
For example, if running a “rpm -qa | grep xCAT” command, without exactmatch set, if the node containes more xCAT rpms that listed in the template, it would be considered a match, as long as all rpms listed in the template were on the node. With exactmatch set, the output must be identical to the template.
--devicetype type_of_device
Specify a user-defined device type that references the location of relevant device configuration file. The devicetype value must correspond to a valid device configuration file. xCAT ships some default configuration files for Ethernet switches and and IB switches under /opt/xcat/share/xcat/devicetype directory. If you want to overwrite any of the configuration files, please copy it to /var/opt/xcat/directory and cutomize it. For example, base/IBSwitch/Qlogic/config is the configuration file location if devicetype is specified as IBSwitch::Qlogic. xCAT will first search config file using /var/opt/xcat/ as the base. If not found, it will search for it using /opt/xcat/share/xcat/devicetype/ as the base.
-l | --user user_ID
Specifies a remote user name to use for remote command execution.
-c | --command
The xdsh or rinv command that will be run. The command should be enclosed in double quotes to insure correct shell interpretation. This parameter must only contain, the node range or the image path (Linux) or spot name for AIX. It cannot be used to set additional input flags to xdsh or rinv (for example -s,-T,-e). See examples below.
Note: do not add the | xdshcoll to the command, it is automatically added by sinv. sinv also automatically sets the -v flag for xdsh.
-f | --file
The file containing the xdsh or rinv command that will be run. This should be the fully qualified name of the file.
Note: do not add the | xdshcoll to the command in the file, it is automatically added by sinv.
-r | --remove
This flag indicates that generated templates should be removed at the at the end of the sinv command execution.
If the flag is input, then all templates that are generated by the sinvcommand, will be removed. If the first template is created by the admin, it will not be removed.
If the flag is not input, no templates will be removed. It is up to the admin to cleanup templates.
-h | --help
Displays usage information.
-v | --version
Displays xCAT release version.
-V | --Verbose
Verbose mode.
*
To setup sinv.template (name optional) for input to the sinv command , enter:
xdsh node1,node2 “rpm -qa | grep ssh ” | xdshcoll > /tmp/sinv.template
- Note: when setting up the template the output of xdsh must be piped
- to xdshcoll, sinv processing depends on it.
*
To setup rinv.template for input to the sinv command , enter:
rinv node1-node2 serial | xdshcoll > /tmp/rinv.template
- Note: when setting up the template the output of rinv must be piped
- to xdshcoll, sinv processing depends on it.
*
To execute sinv using the sinv.template generated above on the nodegroup, testnodes ,possibly generating up to two new templates, and removing all generated templates in the end, and writing output report to /tmp/sinv.output, enter:
sinv * -c “xdsh testnodes rpm -qa | grep ssh” -p /tmp/sinv.template -t 2 -r -o /tmp/sinv.output*
Note: do not add the pipe to xdshcoll on the -c flag, it is automatically added by the sinv routine.
*
To execute sinv on noderange, node1-node4, using the seed node, node8, to generate the first template, using the xdsh command (-c), possibly generating up to two additional templates and not removing any templates at the end, enter:
sinv -c “xdsh node1-node4 lslpp -l | grep bos.adt” -s node8 -p /tmp/sinv.template -t 2 -o /tmp/sinv.output
*
To execute sinv on noderange, node1-node4, using the seed node, node8, to generate the first template, using the rinv command (-c), possibly generating up to two additional templates and removing any generated templates at the end, enter:
sinv -c “rinv node1-node4 serial” -s node8 -p /tmp/sinv.template -t 2 -r -o /tmp/rinv.output
*
To execute sinv on noderange, node1-node4, using node1 as the seed node, to generate the sinv.template from the xdsh command (-c), using the exact match option, generating no additional templates, enter:
sinv -c “xdsh node1-node4 lslpp -l | grep bos.adt” -s node1 -e -p /tmp/sinv.template -o /tmp/sinv.output
Note: the /tmp/sinv.template file must be empty, otherwise it will be used as an admin generated template.
*
To execute sinv on the Linux osimage defined for cn1. First build a template from the /etc/hosts on the node. Then run sinv to compare. xdsh *cn1 “cat /etc/hosts” | xdshcoll * /tmp/sinv2/template”
sinv -c “xdsh -i /install/netboot/rhels6/ppc64/test_ramdisk_statelite/rootimg cat /etc/hosts” -e -t1 -p /tmp/sinv.template -o /tmp/sinv.output
*
*
To execute sinv on the device mswitch2 and compare to mswitch1
sinv -c “xdsh mswitch enable;show version” -s mswitch1 -p /tmp/sinv/template -***-devicetype IBSwitch::Mellanox -l admin -t 2*
Files
/opt/xcat/bin/sinv/
Location of the sinv command.
L <xdsh(1)|xdsh.1>, noderange(3)|noderange.3
snmove.1¶
snmove - Move xCAT compute nodes to a different xCAT service node.
snmove noderange [-V] [-l | --liteonly] [-d | --dest sn2] [-D | --destn sn2n] [-i | --ignorenodes] [-P | --postscripts script1,script2...|all]
snmove [-V] [-l | --liteonly] -s | --source sn1 [-S | --sourcen sn1n] [-d | --dest sn2] [-D | --destn sn2n] [-i | --ignorenodes] [-P | --postscripts script1,script2...|all]
snmove [-h | --help | -v | --version]
The snmove command may be used to move a node or nodes from one service node to another backup service node.
The use of backup service nodes in an xCAT hierarchical cluster can help improve the overall reliability, availability, and serviceability of the cluster.
Before you run the snmove command it is assumed that the backup service node has been configured properly to manage the new node or nodes. (See the xCAT document named “Using xCAT Service Nodes with AIX” for information on how to set up backup AIX service nodes.).
The snmove command can use the information stored in the xCAT database or information passed in on the command line to determine the current service node and the backup service node.
To specify the primary and backup service nodes you can set the “servicenode” attribute of the node definitions.
The servicenode attribute is the hostname of the xCAT service node as it is known by the management node. The xcatmaster attribute is the hostname of the xCAT service node as known by the node. The servicenode attribute should be set to a comma-separated list so that the primary service node is first and the backup service node is second. The xcatmaster attribute must be set to the hostname of the primary service node as it is known by the node.
When the snmove command is run it modifies the xCAT database to switch the the primary server to the backup server.
It will also check the other services that are being used for the node (tftpserver, monserver, nfsserver, conserver), and if they were set to the original service node they will be changed to point to the backup service node.
By default the command will modify the nodes so that they will be able to be managed by the backup service node.
If the -i option is specified, the nodes themselves will not be modified.
You can also have postscripts executed on the nodes by using the -P option if needed.
The xCAT snmove command may also be used to synchronize statelite persistent files from the primary service node to the backup service node without actually moving the nodes to the backup servers.
If you run the command with the “-l” option it will attempt to use rsync to update the statelite persistent directory on the backup service node. This will only be done if the server specified in the “statelite” table is the primary service node.
When the snmove command is executed the new service node must be running but the original service node may be down.
Note: On a Linux cluster, for NFS statelite nodes that do not use external NFS server, if the original service node is down, the nodes it manages will be down too. You must run nodeset command and then reboot the nodes after running snmove. For stateless nodes and RAMDisk statelite nodes, the nodes will be up even if the original service node is down. However, make sure to run nodeset command if you decide to reboot the nodes later.
-d|--dest
Specifies the hostname of the new destination service node as known by (facing) the management node.
-D|--destn
Specifies the hostname of the destination service node as known by (facing) the nodes.
-h|--help
Display usage message.
-i|--ignorenodes
No modifications will be made on the nodes. If not specified, several xCAT postscripts will be run on the nodes to complete the switch to the new service node.
-l|--liteonly
Use this option to ONLY synchronize any AIX statelite files from the primary server to the backup server for the nodes. It will not do the actual moving of thre nodes the the backup servers.
-P|--postscripts
Specifies a list of extra postscripts to be run on the nodes after the nodes are moved over to the new serive node. If ‘all’ is specified, all the postscripts defined in the postscripts table will be run for the nodes. The specified postscripts must be stored under /install/postscripts directory.
-s|--source
- Specifies the hostname of the current (source) service node sa known by (facing)
- the management node.
-S|--sourcen
Specifies the hostname of the current service node adapter as known by (facing) the nodes.
-V|--verbose
Verbose mode.
-v|--version
Command Version.
Move the nodes contained in group “group1” to the service node named “xcatsn02”.
snmove group1 -d xcatsn02 -D xcatsn02-eth1
Move all the nodes that use service node xcatsn01 to service node xcatsn02.
snmove -s xcatsn01 -S xcatsn01-eth1 -d xcatsn02 -D xcatsn02-eth1
Move any nodes that have sn1 as their primary server to the backup service node set in the xCAT node definition.
snmove -s sn1
Move all the nodes in the xCAT group named “nodegroup1” to their backup SNs.
snmove nodegroup1
Move all the nodes in xCAT group “sngroup1” to the service node named “xcatsn2”.
snmove sngroup1 -d xcatsn2
Move all the nodes in xCAT group “sngroup1” to the SN named “xcatsn2” and run extra postscripts.
snmove sngroup1 -d xcatsn2 -P test1
Move all the nodes in xCAT group “sngroup1” to the SN named “xcatsn2” and do not run anything on the nodes.
snmove sngroup1 -d xcatsn2 -i
Synchronize any AIX statelite files from the primary server for compute03 to the backup server. This will not actually move the node to it’s backup service node.
snmove compute03 -l -V
/opt/xcat/sbin/snmove
noderange(3)|noderange.3
swapnodes.1¶
swapnodes - swap the location info in the db (all the attributes in the ppc table and the nodepos table) between 2 nodes. If swapping within a cec, it will assign the IO adapters that were assigned to the defective node to the available node.
This command is only for Power 775 using Direct FSP Management, and used in Power 775 Availability Plus.
The swapnodes command will keep the current_node name in the xCAT table, and use the fip_node‘s hardware resource. Besides that, the IO adapters will be assigned to the new hardware resource if they are in the same CEC. So the swapnodes command will do 2 things:
(1)swap the location info in the db between 2 nodes:
All the ppc table attributes (including hcp, id, parent, supernode and so on).
All the nodepos table attributes(including rack, u, chassis, slot, room and so on).
(2)assign the I/O adapters from the defective node(the original current_node) to the available node(the original fip_node) if the nodes are in the same cec.
The swapnodes command shouldn’t make the decision of which 2 nodes are swapped. It will just received the 2 node names as cmd line parameters.
After running swapnodes command, the order of the I/O devices may be changed after IO re-assignment, so the administrator needs to run rbootseq to set the boot string for the current_node. And then boot the node with the same image and same postscripts because they have the same attributes.
Without -o option, it’s used to swap the location info in the db between 2 nodes. With -o option, it’s used to move the current_node definition to fip_node (the 2nd octant), not move the fip_node definition to the 1st octant. If the two nodes are in a cec, it will assign the IO adapters that were assigned to the defective node to the available node. Originally, the current_node is a defective non-compute node, and fip_node is a avaible compute node. After the swapping, the current_node will be a available node.
-h|--help
Display usage message.
-c
current_node – the defective non-compute node.
-f
fip_node – a compute node which will be swapped as the non-compute node.
-o
one way. Only move the current_node definition to the fip_node‘s hardware resource, and not move the fip_node definition to the current_node. And then the current_node will use the fip_node‘s hardware resource, and the fip_node definition is not changed. if the two nodes are in the same CEC, the I/O adapter from the original current_node will be assigned to the fip_node.
1
To swap the service node attributes and IO assignments between sn1 and compute2 which are in the same cec, all the attributes in the ppc table and nodepos talbe of the two node will be swapped, and the the I/O adapters from the defective node (the original sn1) will be assigned to the available node (the original compute2). After the swapping, the sn1 will use the compute2’s hardware resource and the I/O adapters from the original sn1.
swapnodes -c sn1 -f compute2
2
To swap the service node attributes and IO assignments between sn1 and compute2 which are NOT in the same cec, all the attributes in the ppc table and nodepos talbe of the two node will be swapped. After the swapping, the sn1 will use the compute2’s hardware resource.
swapnodes -c sn1 -f compute2
3
Only to move the service node (sn1) definition to the compute node (compute2)’s hardware resource, and not move the compute2 definition to the sn1. After the swapping, the sn1 will use the compute2’s hardware resource, and the compute2 definition is not changed.
swapnodes -c sn1 -f compute2 -o
$XCATROOT/bin/swapnodes
(The XCATROOT environment variable is set when xCAT is installed. The default value is “/opt/xcat”.)
This command is part of the xCAT software product.
lsvm(1)|lsvm.1, mkvm(1)|mkvm.1, chvm(1)|chvm.1
switchblade.1¶
switchblade MM {list | stat}
switchblade node {media | mt | kvm | video | both} [slot_num]
switchblade [-h | --help | -v | --version]
switchblade assigns the BladeCenter media tray and/or KVM to the specified blade, so that they can be used with that blade. If list or stat are specified instead, switchblade will display the current assignment. You can either specify a management module or a node (blade) to switchblade. If the latter, switchblade will determine the management module of the node.
list | stat
Display which blade the media tray and KVM are currently assigned to.
media | mt
Assign the media tray to the specified blade.
kvm | video
Assign the KVM (video display) to the specified blade.
both
Assign both the media tray and the KVM to the specified blade.
slot_num
The slot # of the blade that the resources should be assigned to. If not specified, it will use the slot # of the node specified.
Switch the media tray to be assigned to the blade in slot 4 (assume it is node4):
switchblade node4 media
Output will be like:
Media Tray slot: 4
rbootseq(1)|rbootseq.1
tabgrep.1¶
tabgrep - list table names in which an entry for the given node appears.
The tabgrep command displays the tables that contain a row for the specified node. Note that the row can either have that nodename as the key or it could have a group that contains the node as the key.
*
To display the tables that contain blade1:
tabgrep blade1
The output would be similar to:
nodelist nodehm mp chain hosts mac noderes nodetype
/opt/xcat/bin/tabgrep
nodels(1)|nodels.1, tabdump(8)|tabdump.8
unregnotif.1¶
unregnotif - unregister a Perl module or a command that was watching for the changes of the desired xCAT database tables.
This command is used to unregistered a Perl module or a command that was watching for the changes of the desired xCAT database tables.
filename is the path name of the Perl module or command to be registered.
-h | -help Display usage message.
**-v | -version ** Command Version.
-V | -verbose Verbose output.
- To unregistered a Perl module, enter:
unregnotif /opt/xcat/lib/perl/xCAT_monitoring/mycode.pm
- To register a command, enter:
unregnotif /usr/bin/mycmd
/opt/xcat/bin/unregnotif
regnotif(1)|regnotif.1
updateSNimage.1¶
updateSNimage - Adds the needed Service Node configuration files to the install image.
This command is used to add the Service Node configuration files to the install image. It will either copy them locally or scp them to a remote host.
-h |--help Display usage message.
-v |--version Display xCAT version.
-n | --node A remote host name or ip address that contains the install image to be updated.
-p |--path Path to the install image.
- To update the image on the local host.
updateSNimage -p /install/netboot/fedora8/x86_64/test/rootimg
- To update the image on a remote host.
updateSNimage -n 9.112.45.6 -p /install/netboot/fedora8/x86_64/test/rootimg
updatenode.1¶
updatenode - Update nodes in an xCAT cluster environment.
updatenode noderange [-V | --verbose] [-F | --sync] [-f | --snsync] [-S | --sw] [-l userID] [-P | --scripts [script1,script2...]] [-s | --sn] [-A | --updateallsw] [-c | --cmdlineonly] [-d alt_source_dir] [--fanout] [-t timeout} [attr=val [attr=val...]]
updatenode noderange [-k | --security] [-t timeout]
updatenode noderange [-g | --genmypost]
updatenode noderange [-V | --verbose] [-t timeout] [script1,script2...]
updatenode noderange [-V | --verbose] [-f | --snsync]
updatenode [-h | --help] [-v | --version]
The updatenode command is run on the xCAT management node and can be used to perform the following node updates:
1
Distribute and synchronize files.
2
Install or update software on diskfull nodes.
3
Run postscripts.
4
Update the ssh keys and host keys for the service nodes and compute nodes; Update the ca and credentials for the service nodes.
The default behavior when no options are input to updatenode will be to run the following options “-S”, “-P” and “-F” options in this order. If you wish to limit updatenode to specific actions you can use combinations of the “-S”, “-P”, and “-F” flags.
For example, If you just want to synchronize configuration file you could specify the “-F” flag. If you want to synchronize files and update software you would specify the “-F” and “-S” flags. See the descriptions of these flags and examples below.
The flag “-k” (–security) can NOT be used together with “-S”, “-P”, and “-F” flags.
The flag “-f” (–snsync) can NOT be used together with “-S”, “-P”, and “-F” flags.
Note: In a large cluster environment the updating of nodes in an ad hoc manner can quickly get out of hand, leaving the system administrator with a very confusing environment to deal with. The updatenode command is designed to encourage users to handle cluster updates in a manner that is recorded and easily repeatable.
The basic process for distributing and synchronizing nodes is:
*
Create a synclist file.
*
Indicate the location of the synclist file.
*
Run the updatenode command to update the nodes.
Files may be distributed and synchronized for both diskless and diskfull nodes. Syncing files to NFS-based statelite nodes is not supported.
- More information on using the synchronization file function is in
- the following doc: Using_Updatenode.
The synclist file contains the configuration entries that specify where the files should be synced to. In the synclist file, each line is an entry which describes the location of the source files and the destination location for the files on the target node.
For more information on creating your synclist files and where to put them, read:
Sync-ing_Config_Files_to_Nodes
updatenode <noderange> -F
updatenode can be use to install or update software on the nodes. See the following documentation for setting up otherpkgs: Install_Additional_Packages
To install/update the packages, run:
updatenode <noderange> -S
For Linux systems:
It this is equivalent to running the following command:
updatenode noderange -P ospkgs,otherpkgs
It will update all the rpms specified in the .pkglist file and .otherpkgs.pkglist file. ospkgs postscript will normally remove all the existing rpm repositories before adding server:/install/<os>/<arch/ as the new repository. To preserve the existing repositories, you can run the following command instead:
updatenode noderange -P “ospkgs –keeprepo,otherpkgs”
For AIX systems:
Note: The updatenode command is used to update AIX diskfull nodes only. For updating diskless AIX nodes refer to the xCAT for AIX update documentation and use the xCAT mknimimage command. For information on updating software on AIX cluster: For diskful installs, read: XCAT_AIX_RTE_Diskfull_Nodes For diskless installs, read: XCAT_AIX_Diskless_Nodes
updatenode can also be used in Sysclone environment to push delta changes to target node. After capturing the delta changes from the golden client to management node, just run below command to push delta changes to target nodes.
updatenode <targetnoderange> -S
The scripts must be copied to the /install/postscripts directory on the xCAT management node. (Make sure they are executable and world readable.)
To run scripts on a node you must either specify them on the command line or you must add them to the “postscripts” attribute for the node.
To set the postscripts attribute of the node (or group) definition you can use the xCAT chdef command. Set the value to be a comma separated list of the scripts that you want to be executed on the nodes. The order of the scripts in the list determines the order in which they will be run. You can use the lsdef command to check the postscript order.
Scripts can be run on both diskless and diskfull nodes.
To run all the customization scripts that have been designated for the nodes, (in the “postscripts and postbootscripts” attributes), type:
updatenode <noderange> -P
To run the “syslog” script for the nodes, type:
updatenode <noderange> -P syslog
To run a list of scripts, type:
updatenode <noderange> -P "script1 p1 p2,script2"
where p1 p2 are the parameters for script1.
The flag ‘-P’ can be omitted when only scripts names are specified.
Note: script1,script2 may or may not be designated as scripts to automatically run on the node. However, if you want script1 and script2 to get invoked next time the nodes are deployed then make sure to add them to the “postscripts/postbootscripts” attribute in the database for the nodes.
The basic functions of update security for nodes:
*
Setup the ssh keys for the target nodes. It enables the management node and service nodes to ssh to the target nodes without password.
*
Redeliver the host keys to the target nodes.
*
Redeliver the ca and certificates files to the service node. These files are used to authenticate the ssl connection between xcatd’s of management node and service node.
*
Remove the entries of target nodes from known_hosts file.
Set up the SSH keys
A password for the user who is running this command is needed to setup the ssh keys. This user must have the same uid and gid as the userid on the target node where the keys will be setup.
If the current user is root, roots public ssh keys will be put in the authorized_keys* files under roots .ssh directory on the node(s). If the current user is non-root, the user must be in the policy table and have credential to run the xdsh command. The non-root users public ssh keys and root’s public ssh keys will be put in the authorized_keys* files under the non-root users .ssh directory on the node(s ).
Handle the hierarchical scenario
When update security files for the node which is served by a service node, the service node will be updated automatically first, and then the target node.
The certificates files are needed for a service node to authenticate the ssl connections between the xCAT client and xcatd on the service node, and the xcatd’s between service node and management node. The files in the directories /etc/xcat/cert/ and ~/.xcat/ will be updated.
Since the certificates have the validity time, the ntp service is recommended to be set up between management node and service node.
- Simply running following command to update the security keys:
- updatenode noderange -k
noderange
A set of comma delimited xCAT node names and/or group names. See the xCAT “noderange” man page for details on additional supported formats.
script1,script2...
A comma-separated list of script names. The scripts must be executable and copied to the /install/postscripts directory. Each script can take zero or more parameters. If parameters are spcified, the whole list needs to be quoted by double quotes. For example:
“script1 p1 p2,script2”
[attr=val [attr=val...]]
Specifies one or more “attribute equals value” pairs, separated by spaces. Attr=val pairs must be specified last on the command line. The currently supported attributes are: “installp_bundle”, “otherpkgs”, “installp_flags”, “emgr_flags” and “rpm_flags”. These attribute are only valid for AIX software maintenance support.
--fanout=fanout_value
Specifies a fanout value for the maximum number of concur- rently executing remote shell processes. Serial execution can be specified by indicating a fanout value of 1. If --fanoutis not specified, a default fanout value of 64 is used.
-A|--updateallsw
Install or update all software contained in the source directory. (AIX only)
-c|cmdlineonly
Specifies that the updatenode command should only use software maintenance information provided on the command line. This flag is only valid when using AIX software maintenance support.
-d alt_source_dir
Used to specify a source directory other than the standard lpp_source directory specified in the xCAT osimage definition. (AIX only)
-F|--sync
Specifies that file synchronization should be performed on the nodes. rsync and ssh must be installed and configured on the nodes. The function is not supported for NFS-based statelite installations. For NFS-based statelite installations to sync files, you should use the read-only option for files/directories listed in litefile table with source location specified in the litetree table.
-f|--snsync
Specifies that file synchronization should be performed to the service nodes that service the nodes in the noderange. This updates the service nodes with the data to sync to the nodes. rsync and ssh must be installed and configured on the service nodes. For hierarchy, this optionally can be done before syncing the files to the nodes with the -F flag. If the -f flag is not used, then the -F flag will sync the servicenodes before the nodes automatically. When installing nodes in a hierarchical cluster, this flag should be used to sync the service nodes before the install, since the files will be sync’d from the service node by the syncfiles postscript during the install. The function is not supported for NFS-based statelite installations. For statelite installations to sync files, you should use the read-only option for files/directories listed in litefile table with source location specified in the litetree table.
-g|--genmypost
Will generate a new mypostscript file for the nodes in the noderange, if site precreatemypostscripts is 1 or YES.
-h|--help
Display usage message.
-k|--security
Update the ssh keys and host keys for the service nodes and compute nodes; Update the ca and credentials to the service nodes. Never run this command to the Management Node, it will take down xcatd. You must be running updatenode as root to use the -k flag.
-l | --user user_ID
Specifies a non-root user name to use for remote command execution. This option is only available when running postscripts (-P) for AIX and Linux and updating software (-S) for Linux only. The non-root userid must be previously defined as an xCAT user. The userid sudo setup will have to be done by the admin on the node. This is not supported in a hiearchical cluster, that is the node is serviced by a service node. See the document Granting_Users_xCAT_privileges for required xcat/sudo setup.
-P|--scripts
Specifies that postscripts and postbootscripts should be run on the nodes. updatenode -P syncfiles is not supported. The syncfiles postscript can only be run during install. You should use updatenode <noderange> -F instead.
-S|--sw
Specifies that node software should be updated. In Sysclone environment, specifies pushing the delta changes to target nodes.
-s|--sn
Set the server information stored on the nodes in /opt/xcat/xcatinfo on Linux.
-t timeout
Specifies a timeout in seconds the command will wait for the remote targets to complete. If timeout is not specified it will wait indefinitely. updatenode -k is the exception that has a timeout of 10 seconds, unless overridden by this flag.
-v|--version
Command Version.
-V|--verbose
Verbose mode.
1
To perform all updatenode features for the Linux nodes in the group “compute”:
updatenode compute
The command will: run any scripts listed in the nodes “postscripts and postbootscripts” attribute, install or update any software indicated in the /install/custom/install/<ostype>/profile.otherpkgs.pkglist (refer to the To install or update software part), synchronize any files indicated by the synclist files specified in the osimage “synclists” attribute.
2
To run postscripts,postbootscripts and file synchronization only on the node “clstrn01”:
updatenode clstrn01 -F -P
3
Running updatenode -P with the syncfiles postscript is not supported. You should use updatenode -F instead.
Do not run:
updatenode clstrno1 -P syncfiles
Run:
updatenode clstrn01 -F
4
To run the postscripts and postbootscripts indicated in the postscripts and postbootscripts attributes on the node “clstrn01”:
updatenode clstrn01 -P
5
To run the postscripts script1 and script2 on the node “clstrn01”:
cp script1,script2 /install/postscripts
updatenode clstrn01 -P “script1 p1 p2,script2”
Since flag ‘-P’ can be omitted when only script names are specified, the following command is equivalent:
updatenode clstrn01 “script1 p1 p2,script2”
p1 p2 are parameters for script1.
6
To synchronize the files on the node “clstrn01”: Prepare the synclist file. For AIX, set the full path of synclist in the osimage table synclists attribute. For Linux, put the synclist file into the location: /install/custom/<inst_type>/<distro>/<profile>.<os>.<arch>.synclist Then:
updatenode clstrn01 -F
7
To perform the software update on the Linux node “clstrn01”: Copy the extra rpm into the /install/post/otherpkgs/<os>/<arch>/* and add the rpm names into the /install/custom/install/<ostype>/profile.otherpkgs.pkglist . Then:
updatenode clstrn01 -S
8
To update the AIX node named “xcatn11” using the “installp_bundle” and/or “otherpkgs” attribute values stored in the xCAT database. Use the default installp, rpm and emgr flags.
updatenode xcatn11 -V -S
Note: The xCAT “xcatn11” node definition points to an xCAT osimage definition which contains the “installp_bundle” and “otherpkgs” attributes as well as the name of the NIM lpp_source resource.
9
To update the AIX node “xcatn11” by installing the “bos.cpr” fileset using the “-agQXY” installp flags. Also display the output of the installp command.
updatenode xcatn11 -V -S otherpkgs=”I:bos.cpr” installp_flags=”-agQXY”
Note: The ‘I:’ prefix is optional but recommended for installp packages.
10
To uninstall the “bos.cpr” fileset that was installed in the previous example.
updatenode xcatn11 -V -S otherpkgs=”I:bos.cpr” installp_flags=”-u”
11
To update the AIX nodes “xcatn11” and “xcatn12” with the “gpfs.base” fileset and the “rsync” rpm using the installp flags “-agQXY” and the rpm flags “-i –nodeps”.
updatenode xcatn11,xcatn12 -V -S otherpkgs=”I:gpfs.base,R:rsync-2.6.2-1.aix5.1.ppc.rpm” installp_flags=”-agQXY” rpm_flags=”-i --nodeps”
Note: Using the “-V” flag with multiple nodes may result in a large amount of output.
12
To uninstall the rsync rpm that was installed in the previous example.
updatenode xcatn11 -V -S otherpkgs=”R:rsync-2.6.2-1” rpm_flags=”-e”
13
Update the AIX node “node01” using the software specified in the NIM “sslbnd” and “sshbnd” installp_bundle resources and the “-agQXY” installp flags.
updatenode node01 -V -S installp_bundle=”sslbnd,sshbnd” installp_flags=”-agQXY”
14
To get a preview of what would happen if you tried to install the “rsct.base” fileset on AIX node “node42”. (You must use the “-V” option to get the full output from the installp command.)
updatenode node42 -V -S otherpkgs=”I:rsct.base” installp_flags=”-apXY”
15
To check what rpm packages are installed on the AIX node “node09”. (You must use the “-c” flag so updatenode does not get a list of packages from the database.)
updatenode node09 -V -c -S rpm_flags=”-qa”
16
To install all software updates contained in the /images directory.
updatenode node27 -V -S -A -d /images
Note: Make sure the directory is exportable and that the permissions are set correctly for all the files. (Including the .toc file in the case of installp filesets.)
17
Install the interim fix package located in the /efixes directory.
updatenode node29 -V -S -d /efixes otherpkgs=E:IZ38930TL0.120304.epkg.Z
18
To uninstall the interim fix that was installed in the previous example.
updatenode xcatsn11 -V -S -c emgr_flags=”-r -L IZ38930TL0”
19
To update the security keys for the node “node01”
updatenode node01 -k
20
To update the service nodes with the files to be synchronized to node group compute:
updatenode compute -f
21
To run updatenode with the non-root userid “user1” that has been setup as an xCAT userid with sudo on node1 to run as root, do the following: See Granting_Users_xCAT_privileges for required sudo setup.
updatenode node1 -l user1 -P syslog
22
In Sysclone environment, after capturing the delta changes from golden client to management node, to run updatenode to push these delta changes to target nodes.
updatenode target-node -S
/opt/xcat/bin/updatenode
wcons.1¶
wcons - windowed remote console
wcons provides access to the remote node serial console of a single or range or nodes or groups.
wcons is a simple front-end to rcons in an xterm session for each console.
-t | --tile=n
Tile wcons windows from top left to bottom right. If n is spec- ified then tile n across. If n is not specified then tile to edge of screen. If tiled wcons windows reach bottom right, then the windows start at top left overlaying existing wcons windows.
-h | --help
Print help.
-v | --version
Print version.
xterm options
See xterm(1). Any options other than those listed above are passed directly to xterm. Note: when given multiple nodes, wcons will override -title and tries to figure out optimal -geometryoptions for the xterms (however, -geometry can still be specified).
nodehm table - xCAT node hardware management table. See nodehm(5)|nodehm.5 for further details. This is used to determine the console access method.
wcons node1-node5
wcons --tile --font=nil2 all
wcons -t 4 node1-node16
wcons -f vs -t 4 node1-node4
Tile mode assumes that the width of the left window border is also the width of the right and bottom window border. Most window managers should not have a problem. If you really need support for a screwy window manager let me know.
noderange(3)|noderange.3, rcons(1)|rcons.1, xterm(1)
wkill.1¶
wkill - kill windowed remote consoles
wkill will kill the wcons windows on your $DISPLAY for a single or range or nodes or groups.
wkill was written because I’m too lazy to point and click off 64 windows.
wkill will only kill windows on your display and for only the noderange(3)|noderange.3 you specify. If no noderange(3)|noderange.3 is specified, then all wcons windows on your $DISPLAY will be killed.
wkill node1-node5
noderange(3)|noderange.3, wcons(1)|wcons.1
wvid.1¶
wvid - windowed remote video console for nodes
wvid noderange
wvid provides access to the remote node video console of a single node, or range of nodes or groups. wvid provides a simple front-end to the hardware’s remote console capability. Currently this command is supported for: blades, BMC/IMM, KVM, and Xen
The nodehm.cons attribute of the node determines the method used to open the console. See nodehm(5)|nodehm.5 for further details.
No options are supported at this time.
To open video consoles for the 1st 2 nodes:
wvid node1,node2
noderange(3)|noderange.3, rcons(1)|rcons.1, wcons(1)|wcons.1
xCATWorld.1¶
xCATWorld - Sample client program for xCAT.
xCATWorld {noderange}
The xCATWorld program gives you a sample client program that interfaces to the /opt/xcat/lib/perl/xCAT_plugin/xCATWorld.pm plugin. For debugging purposes we have an Environment Variable XCATBYPASS. If export XCATBYPASS=yes, the client will call the plugin without going through the xcat daemon, xcatd.
/opt/xcat/bin/xCATWorld
This command is part of the xCAT software product.
xcat2nim.1¶
xcat2nim - Use this command to create and manage AIX NIM definitions based on xCAT node, group and network object definitions.
xcat2nim [-h|–help ]
*xcat2nim [-V|–verbose] [-u|–update] [-l|–list] [-r|–remove] [-f|–force] [-t object-types] [-o object-names] [-a|–allobjects] [-p|–primarySN] [-b|–backupSN] [noderange] [attr=val [attr=val...]] *
The xcat2nim command uses xCAT node, group and network object definitions to create, update, list, or remove corresponding NIM definitions.
Before you create or update NIM definitions the xCAT definitions must be created and NIM must be configured.
The xcat2nim command uses xCAT database information, command line input, and default values to run the appropriate NIM commands.
The xCAT node, group and network definition names will correspond to the NIM machine, machine group and network definitions.
Note: The length of a NIM object name must be no longer than 39 characters.
To create or update a NIM definition you must provide the names of the xCAT definitions to use. The default behavior is to create new NIM definitions but not apply updates to existing definitions. If you wish to update existing NIM definitions then you must use the “update” option. If you wish to completely remove the old definition and re-create it you must use the “force” option.
The xCAT code uses the appropriate NIM commands to create the NIM definitions. To create definitions the “nim -o define” operation is used. To update definitions the “nim -o change” operation is used. If you wish to specify additional information to pass to the NIM commands you can use the “attr=val” support. The attribute names must correspond to the attributes supported by the relevant NIM commands. (For example. “netboot_kernel=mp”)
If the object type you are creating is a node then the object names can be a noderange value.
If you are using xCAT service nodes the xcat2nim command will automatically determine the correct server for the node and create the NIM definitions on that server.
The xcat2nim command support for NIM networks is limited to creating and listing.
When creating network definitions the command will check to make sure the network definition (or it’s equivalent) does not exist and then create the required NIM network, route and interface definitions. In some cases the equivalent network definition may exist using a different name. In this case a new definition WILL NOT be created.
To list the NIM definitions that were created you must specify the “list” option and the names of the xCAT objects that were used to create the NIM definitions. The xcat2nim command will list the corresponding NIM machine, machine group or network definitions using the “lsnim -l” command.
To remove NIM definitions you must specify the “remove” option and the names of the xCAT objects that were used to create the NIM definitions.
The remove(“-r”), force(“-f”) and update(“-u”) options are not supported for NIM network definitions.
-a|--all The list of objects will include all xCAT node, group and network objects.
attr=val [attr=val ...] Specifies one or more “attribute equals value” pairs, separated by spaces. Attr=val pairs must be specified last on the command line. The attribute names must correspond to the attributes supported by the relevant NIM commands. When providing attr=val pairs on the command line you must not specify more than one object type.
-b|--backupSN When using backup service nodes only update the backup. The default is to update both the primary and backup service nodes.
-f|--force The force option will remove the existing NIM definition and create a new one.
-h|--help Display the usage message.
-l|--list List NIM definitions corresponding to xCAT definitions.
-o object-names A set of comma delimited xCAT object names. Objects must be of type node, group, or network.
-p|--primarySN When using backup service nodes only update the primary. The default is to update both the primary and backup service nodes.
-r|--remove Remove NIM definitions corresponding to xCAT definitions.
-t object-types A set of comma delimited xCAT object types. Supported types include: node, group, and network.
Note: If the object type is “group”, it means that the xcat2nim command will operate on a NIM machine group definition corresponding to the xCAT node group definition. Before creating a NIM machine group, all the NIM client nodes definition must have been created.
-u|--update Update existing NIM definitions based on xCAT definitions.
-V|--verbose Verbose mode.
- To create a NIM machine definition corresponding to the xCAT node “clstrn01”.
xcat2nim -t node -o clstrn01
- To create NIM machine definitions for all xCAT node definitions.
xcat2nim -t node
- Update all the NIM machine definitions for the nodes contained in the xCAT “compute” node group and specify attribute values that will be applied to each definition.
xcat2nim -u -t node -o compute netboot_kernel=mp cable_type=”N/A”
- To create a NIM machine group definition corresponding to the xCAT group “compute”.
xcat2nim -t group -o compute
- To create NIM network definitions corresponding to the xCAT “clstr_net” an “publc_net” network definitions. Also display verbose output.
xcat2nim -V -t network -o “clstr_net,publc_net”
- To list the NIM definition for node clstrn02.
xcat2nim -l -t node clstrn02
- To re-create a NIM machine definiton and display verbose output.
xcat2nim -V -t node -f clstrn05
- To remove the NIM definition for the group “AIXnodes”.
xcat2nim -t group -r -o AIXnodes
- To list the NIM “clstr_net” definition.
xcat2nim -l -t network -o clstr_net
$XCATROOT/bin/xcat2nim
This command is part of the xCAT software product.
mkdef(1)|mkdef.1
xcatchroot.1¶
xcatchroot - Use this xCAT command to modify an xCAT AIX diskless operating system image.
For AIX diskless images this command will modify the AIX SPOT resource using the chroot command. You must include the name of an xCAT osimage definition and the command that you wish to have run in the spot.
WARNING:
Be very careful when using this command!!! Make sure you are
very clear about exactly what you are changing so that you do
not accidently corrupt the image.
As a precaution it is advisable to make a copy of the original
spot in case your changes wind up corrupting the image.
When you are done updating a NIM spot resource you should always run the NIM check operation on the spot.
nim -Fo check <spot_name>
The xcatchroot command will take care of any of the required setup so that the command you provide will be able to run in the spot chroot environment. It will also mount the lpp_source resource listed in the osimage definition so that you can access additional software that you may wish to install.
For example, assume that the location of the spot named in the xCAT osimage definition is /install/nim/spot/614spot/usr. The associated root directory in this spot would be /install/nim/spot/614spot/usr/lpp/bos/inst_root. The chroot is automatically done to this new root directory. The spot location is mounted on /.../inst_root/usr so that when your command is run in the chroot environment it is actually running commands from the spot usr location.
Also, the location of the lpp_source resource specified in the osimage definition will be mounted to a subdirectory of the spot /.../inst_root directory. For example, if the lpp_source location is /install/nim/lpp_source/614lpp_lpp_source then that would be mounted over /install/nim/spot/614spot/usr/lpp/bos/inst_root/lpp_source.
When you provide a command string to run make sure you give the full paths of all commands and files assuming the /.../inst_root directory is you root directory.
If you wish to install software from the lpp_source location you would provide a directory location of /lpp_source (or /lpp_source/installp/ppc or /lpp_source/RPMS/ppc etc.) See the example below.
Always run the NIM check operation after you are done updating your spot. (ex. “nim -o check <spot_name>”)
cmd_string
The command you wish to have run in the chroot environment. (Use a quoted string.)
-h |--help
Display usage message.
-i osimage_name
The name of the xCAT osimage definition.
-V |--verbose
Verbose mode.
1) Set the root password to “cluster” in the spot so that when the diskless node boots it will have a root password set.
xcatchroot -i 614spot “/usr/bin/echo root:cluster | /usr/bin/chpasswd -c”
- Install the bash rpm package.
xcatchroot -i 614spot “/usr/bin/rpm -Uvh /lpp_source/RPMS/ppc bash-3.2-1.aix5.2.ppc.rpm”
- To enable system debug.
xcatchroot -i 614spot “bosdebug -D -M”
- To set the “ipforwarding” system tunable.
xcatchroot -i 614spot “/usr/sbin/no -r -o ipforwarding=1”
/opt/xcat/bin/xcatchroot
This command is part of the xCAT software product.
xcattest.1¶
xcattest - Run automated xCAT test cases.
xcattest [-?|-h]
xcattest [-f configure file] [-b case bundle list] [-l]
xcattest [-f configure file] [-t case list]
xcattest [-f configure file] [-c cmd list] [-l]
The xcattest command runs test cases to verify the xCAT functions, it can be used when you want to verify the xCAT functions for whatever reason, for example, to ensure the code changes you made do not break the existing commands; to run acceptance test for new build you got; to verify the xCAT snapshot build or development build before putting it onto your production system. The xcattest command is part of the xCAT package xCAT-test.
The root directory for the xCAT-test package is /opt/xcat/share/xcat/tools/autotest/. All test cases are in the sub directory testcase, indexed by the xCAT command, you can add your own test cases according to the test cases format below. The subdirectory bundle contains all the test cases bundles definition files, you can customize or create any test cases bundle file as required. The testing result information will be written into the subdirectory result, the timestamps are used as the postfixes for all the result files. xCAT-test package ships two configuration files template aix.conf.template and linux.conf.template for AIX and Linux environment, you can use the template files as the start point of making your own configuration file.
-?|-h
Display usage message.
-f configure file
Specifies the configuration file, if not specified, the default configure file is /opt/xcat/share/xcat/tools/autotest/default.conf.
-b case bundle list
Comma separated list of test cases bundle files, each test cases bundle can contain multiple lines and each line for one test case name.
-t cases list
Comma separated list of test cases that will be run.
-c cmd list
Comma separated list of commands which will be tested, i.e., all the test cases under the command sub directory will be run.
-l
Display the test cases names specified by the flag -b, -t or -c.
The xCAT-test test cases are in flat text format, the testing framework will parse the test cases line by line, here is an example of the test case:
#required, case name
start:case name
#optional, description of the test case
description: what the test case is for?
#optional, environment requirements
os:AIX/Linux
#optional, environment requirements
arch:ppc/x86
#optional, environment requirements
hcp:hmc/mm/bmc/fsp
#required, command need to run
cmd:comand
#optional, check return code of last executed command
check:rc == or != return code
#optional, check output of last executed command
check:output== or != or =~ or !~ output check string
end
Note: Each test case can have more than one cmd sections and each cmd section can have more than one check:rc sections and more than one check:output sections, the output check string can include regular expressions.
To run all test cases related command rpower:
xcattest -c /tmp/config -c rpower
To run customized bundle:
xcattest -l > /tmp/custom.bundle Modify custom.bundle xcattest -b custom.bundle
To run specified test cases.
xcattest -t lsdef_t_o_l_z
To add a new case to test chvm. In the example, we assume that the min_mem should not be equal to 16 in the lpar profile of computenode. The case name is chvm_custom. It create a test lpar named testnode firstly, that change the min_mem of the lpar to 16 using chvm, then check if min_mem have changed correctly. At last, the testnode be remove to ensure no garbage produced in the cases.
add a new test case file in /opt/xcat/share/xcat/tools/autotest/chvm edit filename start:chvm_custom hcp:hmc cmd:lsvm $$CN > /tmp/autotest.profile check:rc==0 cmd:mkdef -t node -o testnode mgt=hmc groups=all cmd:mkvm testnode -i $$MaxLparID -l $$CN check:rc==0 cmd:perl -pi -e 's/min_mem=\d+/min_mem=16/g' /tmp/autotest.profile cmd:cat /tmp/autotest.profile|chvm testnode check:rc==0 cmd:lsvm testnode check:output=~min_mem=16 cmd:rmvm testnode cmd:rm -f /tmp/autotest.profile end
The xCAT-test testing framework provides some inline functions. The inline functions can be called in test cases as __FUNCTIONNAME(PARAMTERLIST)__ to get some necessary attributes defined in the configuration file. The inline functions can be used in cmd section and the check:output section.
- GETNODEATTR(nodename, attribute) To get the value of specified node’s attribute
- INC(digit) To get value of digit+1.
For example, to run rscan command against the hardware control point of compute node specified in the configuration file:
rscan __GETNODEATTR($$CN, hcp)__ -z
3. B<GETTABLEVALUE(keyname, key, colname, table)> To get the value of column where keyname == key in specified table.
/opt/xcat/bin/xcattest
xcoll.1¶
xcoll - Formats and consolidates the output of the psh, rinv commands.
xcoll [-n] [-c]
The xcoll command formats and consolidates output from the psh,rinv command. The xcollcommand takes, as input, lines in the following format:
groupname: line of output from remote command, will use group name, if defined
The xcoll command formats the lines as follows and writes them to standard output. Assume that the output from node3 and node4 is identical:
====================================
node1 or nodegroup name
====================================
.
.
lines from psh for node1 with hostnames stripped off
.
.
====================================
node2 or nodegroup name
====================================
.
.
lines from psh for node2 with hostnames stripped off
.
.
====================================
node3, node4 or nodegroup name
====================================
.
.
lines from psh for node 3 with hostnames stripped off
.
.
-c
Display a total nodecount for each set of output.
-n
Display output as nodenames instead of groupnames.
*
To display the results of a command issued on several nodes, in the format used in the Description, enter:
psh node1,node2,node3 cat /etc/passwd | xcoll
psh(1)|psh.1, xdshbak(1)|xdshbak.1 ,xdshcoll(1)|xdshcoll.1
xdcp.1¶
xdcp - Concurrently copies files to or from multiple nodes. In addition, provides an option to use rsync to update the files on the nodes, or to an installation image on the local node.
xdcp noderange [[-f fanout] [-L] [-l userID] [-o node_options] [-p] [-P] [-r node_remote_shell] [-R] [-t timeout] [-T] [-v] [-q] [-X env_list] sourcefile.... targetpath
xdcp noderange [-F rsync input file]
xdcp computenoderange [-s -F rsync input file]
xdcp [-i path to install image] [-F rsync input file]
xdcp [-h | -V | -q]
The xdcp command concurrently copies files to or from remote target nodes. The command issues a remote copy com- mand for each node or device specified. When files are pulled from a target, they are placed into the target_path with the name of the remote node or device appended to the copied source_file name. The /usr/bin/rcp command is the model for syntax and security. If using hierarchy, then xdcp runs on the service node that is servicing the compute node. The file will first be copied to the path defined in the site table, SNsyncfiledir attribute, or the default path /var/xcat/syncfiles on the service node, if the attribute is not defined. The -P flag will not automatically copy the files from the compute node to the Management node, hierarchically. There is a two step process, see -P flag. If the Management Node is target node, it must be defined in the xCAT database with nodetype=mn. When the xdcp command runs the Management Node as the target, it does not use remote commands but uses the local OS copy (cp) command.
REMOTE USER:
A user_ID can be specified for the remote copy command. Remote user specification is identical for the xdcp and xdsh commands. See the xdsh command for more information.
REMOTE COMMAND COPY: The xdcp command uses a configurable remote copy command to execute remote copies on remote targets. Support is explicitly provided for Remote Shell rcp command, the OpenSSH scp command and the /usr/bin/rsync command.
For node targets, the remote copy command is determined by the follow- ing order of precedence:
- The -r flag.
- The /usr/bin/scp command.
COMMAND EXECUTIONSThe maximum number of concurrent remote copy command processes (the fanout) can be specified with the -f flag or the DSH_FANOUT environment variable. The fanout is only restricted by the number of remote shell commands that can be run in parallel. You can experiment with the DSH_FANOUT value on your management server to see if higher values are appropriate.
A timeout value for remote copy command execution can be specified with the -t flag or DSH_TIMEOUT environment variable. If any remote target does not respond within the timeout value, the xdcp command displays an error message and exits.
The -T flag provides diagnostic trace information for dcp command exe- cution. Default settings and the actual remote copy commands that are executed to the remote targets are displayed.
The xdcp command can be executed silently using the -Q flag; no target standard output or standard error is displayed.
sourcefile...
Specifies the complete path for the file to be copied to or from the target. Multiple files can be specified. When used with the -R flag, only a single directory can be specified. When used with the -P flag, only a single file can be specified.
targetpath
If one source_file file, then it specifies the file to copy the source_file file to on the target. If multiple source_file files, it specifies the directory to copy the source_file files to on the target. If the -P flag is specified, the target_path is the local host location for the copied files. The remote file directory structure is recreated under target_path and the remote target name is appended to the copied source_file name in the target_path directory. Note: the targetpath directory must exist.
-f | --fanout fanout_value
Specifies a fanout value for the maximum number of concur- rently executing remote shell processes. Serial execution can be specified by indicating a fanout value of 1. If -fis not specified, a default fanout value of 64 is used.
-F | --File rsync input file
Specifies the path to the file that will be used to build the rsync command. The format of the input file is as follows, each line contains:
<path to source file1> <path to source file2> ... -> < path to destination file/directory>or
<path to source file> -> <path to destination file>or
<path to source file> -> <path to destination directory ( must end in /)>
- For example:
- /etc/password /etc/hosts -> /etc
/tmp/file2 -> /tmp/file2 /tmp/file2 -> /tmp/ /tmp/filex -> /tmp/source/filey /etc/* -> /etc/Running postscripts after files are sync’d to the nodes:
After you define the files to rsync, you can add an EXECUTE: clause in the synclist file. The EXECUTE: clause will list all the postscripts that you would like to run after the files are sync’d to the node. The postscript file must be of the form filename.post, where the <filename> is the <filename> is the from <filename>, reside in the same directory as filename, and be executable. If the file filename is rsync’d to the node, then the filename.postwill automatically be run on the node. If the file filename is not updated on the node, the filename.post will not be run.
Putting the filename.post in the file list to rsync to the node is required for hierarchical clusters. It is optional for non-hierarchical clusters.
Another option is the EXECUTEALWAYS: clause in the synclist file. The EXECUTEALWAYS: will list all the postscripts that you would like to run after the files are sync’d to the nodes. These scripts will run whether or not any files are sync’d to the nodes. The scripts have no special format, but must contain the fully qualified path.
The scripts must be also added to the file list to rsync to the node for hierarchical clusters. It is optional for non-hierarchical clusters.
- For example, your rsynclist file may look like this:
- /tmp/share/file2 -> /tmp/file2 /tmp/share/file2.post -> /tmp/file2.post /tmp/share/file3 -> /tmp/filex /tmp/share/file3.post -> /tmp/file3.post /tmp/myscript -> /tmp/myscript # the below are postscripts EXECUTE: /tmp/share/file2.post /tmp/share/file3.post EXECUTEALWAYS: /tmp/myscript
If /tmp/file2 and /tmp/file3 update /tmp/file2 and /tmp/filex on the node, then the postscripts /tmp/file2.post and /tmp/file3.post are automatically run on the node. /tmp/myscript will always be run on the node.
Another option is the APPEND: clause in the synclist file. The APPEND: clause is used to append the contents of the input file to an existing file on the node. The file to append must already exist on the node and not be part of the synclist that contains the APPEND: clause.
- For example, your rsynclist file may look like this:
- /tmp/share/file2 -> /tmp/file2 /tmp/share/file2.post -> /tmp/file2.post /tmp/share/file3 -> /tmp/filex /tmp/share/file3.post -> /tmp/file3.post /tmp/myscript -> /tmp/myscript # the below are postscripts EXECUTE: /tmp/share/file2.post /tmp/share/file3.post EXECUTEALWAYS: /tmp/myscript APPEND: /etc/myappenddir/appendfile -> /etc/mysetup/setup /etc/myappenddir/appendfile2 -> /etc/mysetup/setup2
When you use the append script, the file (left) of the arrow is appended to the file right of the arrow. In this example, /etc/myappenddir/appendfile is appended to /etc/mysetup/setup file, which must already exist on the node. The /opt/xcat/share/xcat/scripts/xdcpappend.sh is used to accomplish this.
Another option is the MERGE: clause in the synclist file. The MERGE: clause is used to append the contents of the input file to /etc/passwd, /etc/group, or /etc/shadow on a Linux node. It is only supported for those files and only on Linux. You must not use both the APPEND and MERGE funcion for these three files. The processing could end up not creating the file you desire. The MERGE function is the preferred method, becuase APPEND only adds to the file. MERGE will add to the file but also insure there are no duplicate entries.
- For example, your rsynclist file may look like this:
- /tmp/share/file2 -> /tmp/file2 /tmp/share/file2.post -> /tmp/file2.post /tmp/share/file3 -> /tmp/filex /tmp/share/file3.post -> /tmp/file3.post /tmp/myscript -> /tmp/myscript # the below are postscripts EXECUTE: /tmp/share/file2.post /tmp/share/file3.post EXECUTEALWAYS: /tmp/myscript APPEND:
/custom/mypasswd -> /etc/passwd /custom/mygroups -> /etc/group /custom/myshadow -> /etc/shadow
Note: no order can be assumed by the order that the EXECUTE,EXECUTEALWAYS and APPEND clause fall in the synclist file.
For more information on syncing files to node, read Sync-ing_Config_Files_to_Nodes
On Linux rsync always uses ssh remoteshell. On AIX, ssh or rsh is used depending on the site.useSSHonAIX attribute.
-h | --help
Displays usage information.
-i | --rootimg install image
Specifies the path to the install image on the local Linux node.
-o | --node-options node_options
Specifies options to pass to the remote shell command for node targets. The options must be specified within double quotation marks (“”) to distinguish them from xdsh options.
-p | --preserve
Preserves the source file characteristics as implemented by the configured remote copy command.
-P | --pull
Pulls (copies) the files from the targets and places them in the target_path directory on the local host. The target_path must be a directory. Files pulled from remote machines have ._target appended to the file name to distinguish between them. When the -P flag is used with the -R flag, ._target is appended to the directory. Only one file per invocation of the xdcp pull command can be pulled from the specified targets. Hierarchy is not automatically support yet. You must first pull the file to the Service Node and then pull the file to the Management node.
-q | --show-config
Displays the current environment settings for all DSH Utilities commands. This includes the values of all environment variables and settings for all currently installed and valid contexts. Each setting is prefixed with context: to identify the source context of the setting.
-r | --node-rcp node_remote_copy
Specifies the full path of the remote copy command used for remote command execution on node targets.
-R | --recursive recursive
Recursively copies files from a local directory to the remote targets, or when specified with the -P flag, recursively pulls (copies) files from a remote directory to the local host. A single source directory can be specified using the source_file parameter.
-s synch service nodes
Will only sync the files listed in the synclist (-F), to the service nodes for the input compute node list. The files will be placed in the directory defined by the site.SNsyncfiledir attribute, or the default /var/xcat/syncfiles directory.
-t | --timeout timeout
Specifies the time, in seconds, to wait for output from any currently executing remote targets. If no output is available from any target in the specified timeout, xdshdisplays an error and terminates execution for the remote targets that failed to respond. If timeout is not specified, xdsh waits indefinitely to continue processing output from all remote targets. When specified with the -i flag, the user is prompted for an additional timeout interval to wait for output.
-T | --trace
Enables trace mode. The xdcp command prints diagnostic messages to standard output during execution to each target.
-v | --verify
Verifies each target before executing any remote commands on the target. If a target is not responding, execution of remote commands for the target is canceled.
-V | --version
Displays the xdcp command version information.
DSH_ENVIRONMENT
Specifies a file that contains environment variable definitions to export to the target before executing the remote command. This variable is overridden by the -E flag.
DSH_FANOUT
Specifies the fanout value. This variable is overridden by the -f flag.
DSH_NODE_OPTS
Specifies the options to use for the remote shell command with node targets only. This variable is overridden by the -o flag.
DSH_NODE_RCP
Specifies the full path of the remote copy command to use to copy local scripts and local environment configuration files to node targets.
DSH_NODE_RSH
Specifies the full path of the remote shell to use for remote command execution on node targets. This variable is overridden by the -r flag.
DSH_NODEGROUP_PATH
Specifies a colon-separated list of directories that contain node group files for the DSH context. When the -a flag is specified in the DSH context, a list of unique node names is collected from all node group files in the path.
DSH_PATH
Sets the command path to use on the targets. If DSH_PATH is not set, the default path defined in the profile of the remote user_ID is used.
DSH_SYNTAX
Specifies the shell syntax to use on remote targets; ksh or csh. If not specified, the ksh syntax is assumed. This variable is overridden by the -S flag.
DSH_TIMEOUT
Specifies the time, in seconds, to wait for output from each remote target. This variable is overridden by the -tflag.
Exit values for each remote copy command execution are displayed in messages from the xdcp command, if the remote copy command exit value is non-zero. A non-zero return code from a remote copy command indicates that an error was encountered during the remote copy. If a remote copy command encounters an error, execution of the remote copy on that tar- get is bypassed.
The xdcp command exit code is 0, if the xdcp command executed without errors and all remote copy commands finished with exit codes of 0. If internal xdcp errors occur or the remote copy commands do not complete successfully, the xdcp command exit value is greater than 0.
The xdcp command has no security configuration requirements. All remote command security requirements - configuration, authentication, and authorization - are imposed by the underlying remote command configured for xdsh. The command assumes that authentication and authorization is configured between the local host and the remote targets. Interactive password prompting is not supported; an error is displayed and execution is bypassed for a remote target if password prompting occurs, or if either authorization or authentication to the remote target fails. Security configurations as they pertain to the remote environment and remote shell command are userdefined.
*
To copy the /etc/hosts file from all nodes in the cluster to the /tmp/hosts.dir directory on the local host, enter:
xdcp all -P /etc/hosts /tmp/hosts.dir
A suffix specifying the name of the target is appended to each file name. The contents of the /tmp/hosts.dir directory are similar to:
hosts._node1 hosts._node4 hosts._node7 hosts._node2 hosts._node5 hosts._node8 hosts._node3 hosts._node6
*
To copy the directory /var/log/testlogdir from all targets in NodeGroup1 with a fanout of 12, and save each directory on the local host as /var/log._target, enter:
xdcp NodeGroup1 -f 12 -RP /var/log/testlogdir /var/log
*
To copy /localnode/smallfile and /tmp/bigfile to /tmp on node1 using rsync and input -t flag to rsync, enter:
xdcp node1 -r /usr/bin/rsync -o “-t” /localnode/smallfile /tmp/bigfile /tmp
*
To copy the /etc/hosts file from the local host to all the nodes in the cluster, enter:
xdcp all /etc/hosts /etc/hosts
*
To copy all the files in /tmp/testdir from the local host to all the nodes in the cluster, enter:
xdcp all /tmp/testdir/\ /tmp/testdir*
*
To copy all the files in /tmp/testdir and it’s subdirectories from the local host to node1 in the cluster, enter:
xdcp node1 -R /tmp/testdir /tmp/testdir
*
To copy the /etc/hosts file from node1 and node2 to the /tmp/hosts.dir directory on the local host, enter:
xdcp node1,node2 -P /etc/hosts /tmp/hosts.dir
*
To rsync the /etc/hosts file to your compute nodes:
Create a rsync file /tmp/myrsync, with this line:
/etc/hosts -> /etc/hosts
or
/etc/hosts -> /etc/ (last / is required)
Run:
xdcp compute -F /tmp/myrsync
*
To rsync all the files in /home/mikev to the compute nodes:
Create a rsync file /tmp/myrsync, with this line:
/home/mikev/* -> /home/mikev/ (last / is required)
Run:
xdcp compute -F /tmp/myrsync
*
To rsync to the compute nodes, using service nodes, the command will first rsync the files to the /var/xcat/syncfiles directory on the service nodes and then rsync the files from that directory to the compute nodes. The /var/xcat/syncfiles default directory on the service nodes, can be changed by putting a directory value in the site table SNsyncfiledir attribute.
Create a rsync file /tmp/myrsync, with this line:
/etc/hosts /etc/passwd -> /etc
or
/etc/hosts /etc/passwd -> /etc/
Run:
xdcp compute -F /tmp/myrsync to update the Compute Nodes
*
To rsync to the service nodes in preparation for rsyncing the compute nodes during an install from the service node.
Create a rsync file /tmp/myrsync, with this line:
/etc/hosts /etc/passwd -> /etc
Run:
xdcp compute -s -F /tmp/myrsync to sync the service node for compute
*
To rsync the /etc/file1 and file2 to your compute nodes and rename to filex and filey:
Create a rsync file /tmp/myrsync, with these line:
/etc/file1 -> /etc/filex
/etc/file2 -> /etc/filey
Run:
xdcp compute -F /tmp/myrsync to update the Compute Nodes
*
To rsync files in the Linux image at /install/netboot/fedora9/x86_64/compute/rootimg on the MN:
Create a rsync file /tmp/myrsync, with this line:
/etc/hosts /etc/passwd -> /etc
Run:
xdcp -i /install/netboot/fedora9/x86_64/compute/rootimg -F /tmp/myrsync
*
To define the Management Node in the database so you can use xdcp,run
xcatconfig -m
xdsh(1)|xdsh.1, noderange(3)|noderange.3
xdsh.1¶
xdsh - Concurrently runs remote commands on multiple nodes (Management Node, Service Nodes, compute nodes), or an install image.
xdsh noderange [-B bypass] [--devicetype type_of_device] [-e] [-E environment_file] [-f fanout] [-L] [-l userID] [-m] [-onode_options] [-Q] [-r node_remote_shell] [-s] [-S csh | ksh] [-t timeout] [-T] [-v] [-X env_list] [-z] [--sudo] command_list
xdsh noderange [-K]
xdsh noderange [-K] [-l userID] --devicetype type_of_device
xdsh [-i image path | nim image name] command_list
xdsh noderange [-c]
xdsh [-h | -V | -q]
The xdsh command runs commands in parallel on remote nodes and/or the Management Node. The xdsh command issues a remote shell command for each target specified, and returns the output from all targets, formatted so that command results from all nodes can be managed. If the command is to be executed on the Management Node, it does not use a remote shell command, but uses the local OS copy or shell command. The Management Node must be defined in the xCAT database. The best way to do this is to use the xcatconfig -m option. The xdsh command is an xCAT Distributed Shell Utility.
COMMAND SPECIFICATION:
The commands to execute on the targets are specified by the command_list xdsh parameter, or executing a local script using the -eflag.
The syntax for the command_list xdsh parameter is as follows:
command[; command]...
where command is the command to run on the remote target. Quotation marks are required to ensure that all commands in the list are executed remotely, and that any special characters are interpreted correctly on the remote target. A script file on the local host can be executed on each of the remote targets by using the -eflag. If -e is specified, command_list is the script name and arguments to the script. For example:
xdsh hostname -e script_filename [arguments]...
The script_filename file is copied to a random filename in the /tmpdirectory on each remote target and then executed on the targets.
The xdsh command does not work with any interactive commands, including those that read from standard input.
REMOTE SHELL COMMAND:
The xdsh command uses a configurable remote shell command to execute remote commands on the remote targets. Support is explicitly provided for AIX Remote Shell and OpenSSH, but any secure remote command that conforms to the IETF (Internet Engineering Task Force) Secure Remote Command Protocol can be used.
The remote shell is determined as follows, in order of precedence:
- The -r flag.
- The DSH_NODE_RSH environment variable.
- The default node remote shell as defined by the target context.
- The /usr/bin/ssh command.
The remote shell options are determined as follows, in order of prece- dence:
- The -o flag.
- The DSH_NODE_OPTS environment variable.
REMOTE SHELL ENVIRONMENT:
The shell environment used on the remote target defaults to the shell defined for the user_ID on the remote target. The command syntax that xdsh uses to form the remote commands can be specified using the -Sflag. If -S is not specified, the syntax defaults to sh syntax.
When commands are executed on the remote target, the path used is determined by the DSH_PATH environment variable defined in the shell of the current user. If DSH_PATH is not set, the path used is the remote shell default path. For example, to set the local path for the remote targets, use:
DSH_PATH=$PATH
The -E flag exports a local environment definition file to each remote target. Environment variables specified in this file are defined in the remote shell environment before the command_list is executed. The definition file should contain entries like the following
- and be executable. One environment variable per line.
- export NEWENVVARIABLE=”yes” export ANOTHERENVVARIABLE=”yes”
COMMAND EXECUTION:
The maximum number of concurrent remote shell command processes (the fanout) can be specified with the -f flag or with the DSH_FANOUTenvironment variable. The fanout is only restricted by the number of remote shell commands that can be run in parallel. You can experiment with the DSH_FANOUT value on your management server to see if higher values are appropriate.
A timeout value for remote command execution can be specified with the -t flag or with the DSH_TIMEOUT environment variable. If any remote target does not provide output to either standard output or standard error within the timeout value, xdsh displays an error message and exits.
If streaming mode is specified with the -s flag, output is returned as it becomes available from each target, instead of waiting for the command_list to complete on all targets before returning output. This can improve performance but causes the output to be unsorted.
The -z flag displays the exit code from the last command issued on the remote node in command_list. Note that OpenSSH behaves differently; it returns the exit status of the last remote command issued as its exit status. If the command issued on the remote node is run in the background, the exit status is not displayed.
The -m flag monitors execution of the xdsh command by printing status messages to standard output. Each status message is preceded by dsh.
The -T flag provides diagnostic trace information for the execution of the xdsh command. Default settings and the actual remote shell commands executed on the remote targets are displayed.
No error detection or recovery mechanism is provided for remote targets. The xdsh command output to standard error and standard output can be analyzed to determine the appropriate course of action.
COMMAND OUTPUT:
The xdsh command waits until complete output is available from each remote shell process and then displays that output before initiating new remote shell processes. This default behavior is overridden by the -s flag.
The xdsh command output consists of standard error and standard output from the remote commands. The xdsh standard output is the standard output from the remote shell command. The xdsh standard error is the standard error from the remote shell command. Each line is prefixed with the host name of the node that produced the output. The host name is followed by the : character and a command output line. A filter for displaying identical outputs grouped by node is provided separately. See the xdshbak command for more information.
A command can be run silently using the -Q flag; no output from each target’s standard output or standard error is displayed.
SIGNALS:
Signal 2 (INT), Signal 3 (QUIT), and Signal 15 (TERM) are propagated to the commands executing on the remote targets.
Signal 19 (CONT), Signal 17 (STOP), and Signal 18 (TSTP) default to xdsh; the xdsh command responds normally to these signals, but the signals do not have an effect on remotely executing commands. Other signals are caught by xdsh and have their default effects on the xdshcommand; all current child processes, through propagation to remotely running commands, are terminated (SIGTERM).
-c | --cleanup
This flag will have xdsh remove all files from the subdirectories of the the directory on the servicenodes, where xdcp stages the copy to the compute nodes as defined in the site table SNsyncfiledir and nodesyncfiledir
attribute, when the target is a service node.It can also be used to remove the nodesyncfiledir directory on the compute nodes, which keeps the backup copies of files for the xdcp APPEND function support, if a compute node is the target.
-e | --execute
Indicates that command_list specifies a local script filename and arguments to be executed on the remote targets. The script file is copied to the remote targets and then remotely executed with the given arguments. The DSH_NODE_RCP environment variables specify the remote copy command to use to copy the script file to node targets.
-E | --environment environment_file
Specifies that the environment_file contains environment variable definitions to export to the target before executing the command_list.
--devicetype type_of_device
Specify a user-defined device type that references the location of relevant device configuration file. The devicetype value must correspond to a valid device configuration file. xCAT ships some default configuration files for Ethernet switches and and IB switches under /opt/xcat/share/xcat/devicetype directory. If you want to overwrite any of the configuration files, please copy it to /var/opt/xcat/directory and cutomize it. For example, base/IBSwitch/Qlogic/config is the configuration file location if devicetype is specified as IBSwitch::Qlogic. xCAT will first search config file using /var/opt/xcat/ as the base. If not found, it will search for it using /opt/xcat/share/xcat/devicetype/ as the base.
-f | --fanout fanout_value
Specifies a fanout value for the maximum number of concur- rently executing remote shell processes. Serial execution can be specified by indicating a fanout value of 1. If -fis not specified, a default fanout value of 64 is used.
-h | --help
Displays usage information.
-i | --rootimg install image
For Linux, Specifies the path to the install image on the local node. For AIX, specifies the name of the osimage on the local node. Run lsnim for valid names. xdsh will chroot (xcatchroot for AIX) to this path and run the xdsh command against the install image. No other xdsh flags, environment variables apply with this input. A noderange is not accepted. Only runs on the local host, normally the Management Node. The command you run must not prompt for input, the prompt will not be returned to you, and it will appear that xdsh hangs.
-K | --ssh-setup
-K | --ssh-setup -l | --user user_ID --devicetype type_of_device
Set up the SSH keys for the user running the command to the specified node list. The userid must have the same uid, gid and password as the userid on the node
where the keys will be setup.If the current user is root, roots public ssh keys will be put in the authorized_keys* files under roots .ssh directory on the node(s). If the current user is non-root, the user must be in the policy table and have credential to run the xdsh command. The non-root users public ssh keys and root’s public ssh keys will be put in the authorized_keys* files under the non-root users .ssh directory on the node(s). Other device types, such as IB switch, are also supported. The device should be defined as a node and nodetype should be defined as switch before connecting. The xdsh -K command must be run from the Management Node.
-l | --user user_ID
Specifies a remote user name to use for remote command exe- cution.
-L | --no-locale
Specifies to not export the locale definitions of the local host to the remote targets. Local host locale definitions are exported by default to each remote target.
-m | --monitor
Monitors remote shell execution by displaying status messages during execution on each target.
-o | --node-options node_options
Specifies options to pass to the remote shell command for node targets. The options must be specified within double quotation marks (“”) to distinguish them from xdsh options.
-q | --show-config
Displays the current environment settings for all DSH Utilities commands. This includes the values of all environment variables and settings for all currently installed and valid contexts. Each setting is prefixed with context: to identify the source context of the setting.
-Q | --silent
Specifies silent mode. No target output is written to stan- dard output or standard error. Monitoring messages are written to standard output.
-r | --node-rsh node_remote_shell
Specifies the path of the remote shell command used for remote command execution on node targets.
-s | --stream
Specifies that output is returned as it becomes available from each target, instead of waiting for the command_listto be completed on a target before returning output.
-S | --syntax csh | ksh
Specifies the shell syntax to be used on the remote target. If not specified, the ksh syntax is used.
--sudo | --sudo
Adding the –sudo flag to the xdsh command will have xdsh run sudo before running the command. This is particular useful when using the -e option. This is required when you input -l with a non-root user id and want that id to be able to run as root on the node. The non-root userid will must be previously defined as an xCAT user, see process for defining non-root ids in xCAT and setting up for using xdsh. The userid sudo setup will have to be done by the admin on the node. This includes, allowing all commands that you would like to run with xdsh by using visudo to edit the /etc/sudoers file. You must disabl ssh tty requirements by commenting out or removing this line in the /etc/sudoes file “#Defaults requiretty”. See the document Granting_Users_xCAT_privileges for sudo setup requirements. This is not supported in a hierarical cluster, that is the nodes are serviced by servicenodes.
-t | --timeout timeout
Specifies the time, in seconds, to wait for output from any currently executing remote targets. If no output is available from any target in the specified timeout, xdshdisplays an error and terminates execution for the remote targets that failed to respond. If timeout is not specified, xdsh waits indefinitely to continue processing output from all remote targets. The exception is the -K flag which defaults to 10 seconds.
-T | --trace
Enables trace mode. The xdsh command prints diagnostic messages to standard output during execution to each target.
-v | --verify
Verifies each target before executing any remote commands on the target. If a target is not responding, execution of remote commands for the target is canceled. When specified with the -i flag, the user is prompted to retry the verification request.
-V | --version
Displays the xdsh command version information.
-X env_list
Ignore xdsh environment variables. This option can take an argument which is a comma separated list of environment variable names that should NOT be ignored. If there is no argument to this option, or the argument is an empty string, all xdsh environment variables will be ignored. This option is useful when running xdsh from within other scripts when you don’t want the user’s environment affecting the behavior of xdsh.
-z | --exit-status
Displays the exit status for the last remotely executed non-asynchronous command on each target. If the command issued on the remote node is run in the background, the exit status is not displayed.
Exit values for each remote shell execution are displayed in messages from the xdsh command, if the remote shell exit values are non-zero. A non-zero return code from a remote shell indicates that an error was encountered in the remote shell. This return code is unrelated to the exit code of the remotely issued command. If a remote shell encounters an error, execution of the remote command on that target is bypassed.
The xdsh command exit code is 0 if the command executed without errors and all remote shell commands finished with exit codes of 0. If internal xdsh errors occur or the remote shell commands do not complete successfully, the xdsh command exit value is greater than 0. The exit value is increased by 1 for each successive instance of an unsuccessful remote command execution. If the remotely issued command is run in the background, the exit code of the remotely issued command is 0.
DEVICETYPE
Specify a user-defined device type. See --devicetype flag.
DSH_ENVIRONMENT
Specifies a file that contains environment variable definitions to export to the target before executing the remote command. This variable is overridden by the -E flag.
DSH_FANOUT
Specifies the fanout value. This variable is overridden by the -f flag.
DSH_NODE_OPTS
Specifies the options to use for the remote shell command with node targets only. This variable is overridden by the -o flag.
DSH_NODE_RCP
Specifies the full path of the remote copy command to use to copy local scripts and local environment configuration files to node targets.
DSH_NODE_RSH
Specifies the full path of the remote shell to use for remote command execution on node targets. This variable is overridden by the -r flag.
DSH_PATH
Sets the command path to use on the targets. If DSH_PATH is not set, the default path defined in the profile of the remote user_ID is used.
DSH_REMOTE_PASSWORD
If DSH_REMOTE_PASSWORD is set to the password of the userid (usually root) that will ssh to the node, then when you use the -K flag, you will not be prompted for a password.
DSH_SYNTAX
Specifies the shell syntax to use on remote targets; ksh or csh. If not specified, the ksh syntax is assumed. This variable is overridden by the -S flag.
DSH_TIMEOUT
Specifies the time, in seconds, to wait for output from each remote target. This variable is overridden by the -tflag.
To provide backward compatibility for scripts written using dsh in AIX and CSM, a tool has been provide groupfiles4dsh, which will build node group files from the xCAT database that can be used by dsh. See man groupfiles4dsh.
The xdsh command has no security configuration requirements. All remote command security requirements - configuration, authentication, and authorization - are imposed by the underlying remote command configured for xdsh. The command assumes that authentication and authorization is configured between the local host and the remote targets. Interactive password prompting is not supported; an error is displayed and execution is bypassed for a remote target if password prompting occurs, or if either authorization or authentication to the remote target fails. Security configurations as they pertain to the remote environment and remote shell command are userdefined.
The dsh command exit code is 0 if the command executed without errors and all remote shell commands finished with exit codes of 0. If internal dsh errors occur or the remote shell commands do not complete successfully, the dsh command exit value is greater than 0. The exit value is increased by 1 for each successive instance of an unsuccessful remote command execution. If the remotely issued command is run in the background, the exit code of the remotely issued command is 0.
*
To set up the SSH keys for root on node1, run as root:
xdsh node1 -K
*
To run the ps -ef ** command on node targets **node1 and node2, enter:
xdsh node1,node2 “ps -ef”
*
To run the ps command on node targets node1 and run the remote command with the -v and -t flag, enter:
xdsh node1,node2 -o”-v -t” ps=item *
To execute the commands contained in myfile in the XCATcontext on several node targets, with a fanout of 1, enter:
xdsh node1,node2 -f 1 -e myfile
*
To run the ps command on node1 and ignore all the dsh environment variable except the DSH_NODE_OPTS, enter:
xdsh node1 -X \`DSH_NODE_OPTS’ ps
*
To run on Linux, the xdsh command “rpm -qa | grep xCAT” on the service node fedora9 diskless image, enter:
xdsh -i /install/netboot/fedora9/x86_64/service/rootimg “rpm -qa | grep xCAT”
*
To run on AIX, the xdsh command “lslpp -l | grep bos” on the NIM 611dskls spot, enter:
xdsh -i 611dskls “/usr/bin/lslpp -l | grep bos”
*
*
To define the QLogic IB switch as a node and to set up the SSH keys for IB switch qswitch with device configuration file /var/opt/xcat/IBSwitch/Qlogic/config and user name username, Enter
chdef -t node -o qswitch groups=all nodetype=switch
xdsh qswitch -K -l username -***-devicetype IBSwitch::Qlogic*
*
To define the Management Node in the database so you can use xdsh, Enter
xcatconfig -m
*
*
To define a BNT Ethernet switch as a node and run a command to create a new vlan with vlan id 3 on the switch.
chdef myswitch groups=all
- tabch switch=myswitch switches.sshusername=admin switches.sshpassword=passw0rd switches.protocol=[ssh|telnet]
- where admin and passw0rd are the SSH user name and password for the switch. If it is for Telnet, add tn: in front of the user name: tn:admin.
<xdsh> myswitch –devicetype EthSwitch::BNT ‘enable;configure terminal;vlan 3;end;show vlan’
*
xdshbak(1)|xdshbak.1, noderange(3)|noderange.3, groupfiles4dsh(1)|groupfiles4dsh.1
xdshbak.1¶
xdshbak - Formats the output of the xdsh command.
xdshbak [-c | -x | -h | -q]
The xdshbak command formats output from the xdsh command. The xdshbakcommand takes, as input, lines in the following format:
host_name: line of output from remote command
The xdshbak command formats the lines as follows and writes them to standard output. Assume that the output from node3 and node4 is identical, and the -c (collapse) flag was specified:
HOSTS --------------------------------------------------------
node1
--------------------------------------------------------------
.
.
lines from xdsh for node1 with hostnames stripped off
.
.
HOSTS --------------------------------------------------------
node2
--------------------------------------------------------------
.
.
lines from xdsh for node2 with hostnames stripped off
.
.
HOSTS --------------------------------------------------------
node3, node4
--------------------------------------------------------------
.
.
lines from xdsh for node 3 with hostnames stripped off
.
.
When output is displayed from more than one node in collapsed form, the host names are displayed alphabetically. When output is not collapsed, output is displayed sorted alphabetically by host name.
If the -q quiet flag is not set then xdshbakcommand writes ”.” for each 1000 lines of output processed (to show progress), since it won’t display the output until it has processed all of it.
If the -x flag is specified, the extra header lines that xdshbak normally displays for each node will be omitted, and the hostname at the beginning of each line is not stripped off, but xdshbak still sorts the output by hostname for easier viewing:
node1: lines from xdsh for node1
.
.
node2: lines from xdsh for node2
.
.
When the xdshbak filter is used and standard error messages are generated, all error messages on standard error appear before all standard output messages. This is true with and without the -c flag.
-c
If the output from multiple nodes is identical it will be collapsed and displayed only once.
-x
Omit the extra header lines that xdshbak normally displays for each node. This provides more compact output, but xdshbak still sorts the output by node name for easier viewing. This option should not be used with -c.
-h
Displays usage information.
-q
Quiet mode, do not display ”.” for each 1000 lines of output.
*
To display the results of a command issued on several nodes, in the format used in the Description, enter:
xdsh node1,node2,node3 cat /etc/passwd | xdshbak
*
To display the results of a command issued on several nodes with identical output displayed only once, enter:
xdsh host1,host2,host3 pwd | xdshbak -c
*
To display the results of a command issued on several nodes with compact output and be sorted alphabetically by host name, enter:
xdsh host1,host2,host3 date | xdshbak -x
xdsh(1)|xdsh.1, xcoll(1)|xcoll.1
xdshcoll.1¶
xdshcoll - Formats and consolidates the output of the xdsh,sinv commands.
xdshcoll
The xdshcoll command formats and consolidates output from the xdsh,sinv command. The xdshcollcommand takes, as input, lines in the following format:
host_name: line of output from remote command
The xdshcoll command formats the lines as follows and writes them to standard output. Assume that the output from node3 and node4 is identical:
====================================
node1
====================================
.
.
lines from xdsh for node1 with hostnames stripped off
.
.
====================================
node2
====================================
.
.
lines from xdsh for node2 with hostnames stripped off
.
.
====================================
node3, node4
====================================
.
.
lines from xdsh for node 3 with hostnames stripped off
.
.
*
To display the results of a command issued on several nodes, in the format used in the Description, enter:
xdsh node1,node2,node3 cat /etc/passwd | xdshcoll
xdshbak(1)|xdshbak.1
xpbsnodes.1¶
xpbsnodes - PBS pbsnodes front-end for a noderange.
xpbsnodes [{noderange}] [{offline | clear | stat | state}]
xpbsnodes [-h | --help] [-v | --version]
-h Display usage message.
-v Command Version.
offline|off Take nodes offline.
clear|online|on Take nodes online.
stat|state Display PBS node state.
/opt/torque/x86_64/bin/xpbsnodes
noderange(3)|noderange.3
man3¶
noderange.3¶
noderange - syntax for compactly expressing a list of node names
Examples:
node1,node2,node8,node20,group1
node14-node56,node70-node203,group1-group10
node1,node2,node8,node20,node14-node56,node70-node203
node[14-56]
f[1-3]n[1-20]
all,-node129-node256,-frame01-frame03
/node.*
^/tmp/nodes
node10+5
10-15,-13
group1@group2
table.attribute<operator>value
noderange is a syntax that can be used in most xCAT commands to conveniently specify a list of nodes. The result is that the command will be applied to a range of nodes, often in parallel.
noderange is a comma-separated list. Each token (text between commas) in the list can be any of the forms listed below:
Individual node or group:
node01
group1
A range of nodes or groups:
node01-node10 (equivalent to: node01,node02,node03,...node10)
node[01-10] (same as above)
node01:node10 (same as above)
node[01:10] (same as above)
f[1-2]n[1-3] (equivalent to: f1n1,f1n2,f1n3,f2n1,f2n2,f2n3)
group1-group3 (equivalent to: group1,group2,group3)
(all the permutations supported above for nodes are also supported for groups)
nodeRange tries to be intelligent about detecting padding, so you can specify “node001-node200” and it will add the proper number of zeroes to make all numbers 3 digits.
An incremented range of nodes:
node10+3 (equivalent to: node10,node11,node12,node13)
A node shorthand range of nodes:
10-20 (equivalent to: node10,node11,node12,...node20)
10+3 (equivalent to: node10,node11,node12,node13)
Currently, the prefix that will be prepended for the above syntax is always “node”. Eventually, the prefix and optional suffix will be settable via the environment variables XCAT_NODE_PREFIX and XCAT_NODE_SUFFIX, but currently this only works in bypass mode.
A regular expression match of nodes or groups:
/node[345].* (will match any nodes that start with node3, node4, or node5)
/group[12].* (will match any groups that start with group1 or group2)
The path of a file containing noderanges of nodes or groups:
^/tmp/nodelist
where /tmp/nodelist can contain entries like:
#my node list (this line ignored)
^/tmp/foo #ignored
node01 #node comment
node02
node03
node10-node20
/group[456].*
-node50
Node ranges can contain any combination:
node01-node30,node40,^/tmp/nodes,/node[13].*,2-10,node50+5
Any individual noderange may be prefixed with an exclusion operator (default -) with the exception of the file operator (default ^). This will cause that individual noderange to be subtracted from the total resulting list of nodes.
The intersection operator @ calculates the intersection of the left and right sides:
group1@group2 (will result in the list of nodes that group1 and group2 have in common)
Any combination or multiple combinations of inclusive and exclusive ranges of nodes and groups is legal. There is no precedence implied in the order of the arguments. Exclusive ranges have precedence over inclusive. Parentheses can be used to explicitly specify precendence of any operators.
Nodes have precedence over groups. If a node range match is made then no group range match will be attempted.
All node and group names are validated against the nodelist table. Invalid names are ignored and return nothing.
Throughout this man page the term xCAT Node Name Format is used. xCAT Node Name Format is defined by the following regex:
^([A-Za-z-]+)([0-9]+)(([A-Za-z-]+[A-Za-z0-9-]*)*)
In plain English, a node or group name is in xCAT Node Name Format if starting from the begining there are:
*
one or more alpha characters of any case and any number of “-” in any combination
*
followed by one or more numbers
*
then optionally followed by one alpha character of any case or “-“
*
followed by any combination of case mixed alphanumerics and “-“
noderange supports node/group names in any format. xCAT Node Name Format is not required, however some node range methods used to determine range will not be used for non-conformant names.
Example of xCAT Node Name Format node/group names:
NODENAME PREFIX NUMBER SUFFIX
node1 node 1
node001 node 001
node-001 node- 001
node-foo-001-bar node-foo- 001 -bar
node-foo-1bar node-foo- 1 bar
foo1bar2 foo 1 bar2
rack01unit34 rack 01 unit34
unit34rack01 unit 34 rack01
pos0134 pos 0134
Generates a list of all nodes (assuming all is a group) listed in the nodelist table less node5 through node10:
all,-node5-node10
Generates a list of nodes 1 through 10 less nodes 3,4,5. Note that node4 is listed twice, first in the range and then at the end. Because exclusion has precedence node4 will be excluded.
node1-node10,-node3-node5,node4
Generates a list of nodes 1 through 10 less nodes 3 and 5.
node1-node10,-node3,-node5
Generates a list of all (assuming `all’ is a group) nodes in the nodelist table less 17 through 32.
-node17-node32,all
Generates a list of nodes 1 through 128, and user nodes 1 through 4.
node1-node128,user1-user4
Generates a list of all nodes (assuming `all’ is a group), less nodes in groups rack1 through rack3 (assuming groups rack1, rack2, and rack3 are defined), less nodes 100 through 200, less nodes in the storage group. Note that node150 is listed but is excluded.
all,-rack1-rack3,-node100-node200,node150,-storage
Generates a list of nodes matching the regex node[23].\*. That is all nodes that start with node2 or node3 and end in anything or nothing. E.g. node2, node3, node20, node30, node21234 all match.
/node[23].*
Generates a list of nodes which have the value hmc in the nodehm.cons attribute.
nodehm.cons==hmc nodehm.cons=~hmc
Generate a list of nodes in the 1st two frames:
f[1-2]n[1-42]
nodels(1)|nodels.1
man5¶
auditlog.5¶
auditlog - a table in the xCAT database.
auditlog Attributes: recid, audittime, userid, clientname, clienttype, command, noderange, args, status, comments, disable
Audit Data log.
recid
The record id.
audittime
The timestamp for the audit entry.
userid
The user running the command.
clientname
The client machine, where the command originated.
clienttype
Type of command: cli,java,webui,other.
command
Command executed.
noderange
The noderange on which the command was run.
args
The command argument list.
status
Allowed or Denied.
comments
Any user-provided notes.
disable
Do not use. tabprune will not work if set to yes or 1
nodels(1), chtab(8), tabdump(8), tabedit(8)
bootparams.5¶
bootparams - a table in the xCAT database.
bootparams Attributes: node, kernel, initrd, kcmdline, addkcmdline, dhcpstatements, adddhcpstatements, comments, disable
Current boot settings to be sent to systems attempting network boot for deployment, stateless, or other reasons. Mostly automatically manipulated by xCAT.
node
The node or group name
kernel
The kernel that network boot actions should currently acquire and use. Note this could be a chained boot loader such as memdisk or a non-linux boot loader
initrd
The initial ramdisk image that network boot actions should use (could be a DOS floppy or hard drive image if using memdisk as kernel)
kcmdline
Arguments to be passed to the kernel
addkcmdline
User specified one or more parameters to be passed to the kernel. For the kernel options need to be persistent after installation, specify them with prefix “R::”
dhcpstatements
xCAT manipulated custom dhcp statements (not intended for user manipulation)
adddhcpstatements
Custom dhcp statements for administrator use (not implemneted yet)
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
boottarget.5¶
boottarget - a table in the xCAT database.
boottarget Attributes: bprofile, kernel, initrd, kcmdline, comments, disable
Specify non-standard initrd, kernel, and parameters that should be used for a given profile.
bprofile
All nodes with a nodetype.profile value equal to this value and nodetype.os set to “boottarget”, will use the associated kernel, initrd, and kcmdline.
kernel
The kernel that network boot actions should currently acquire and use. Note this could be a chained boot loader such as memdisk or a non-linux boot loader
initrd
The initial ramdisk image that network boot actions should use (could be a DOS floppy or hard drive image if using memdisk as kernel)
kcmdline
Arguments to be passed to the kernel
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
cfgmgt.5¶
cfgmgt - a table in the xCAT database.
cfgmgt Attributes: node, cfgmgr, cfgserver, roles, comments, disable
Configuration management data for nodes used by non-xCAT osimage management services to install and configure software on a node.
node
The node being managed by the cfgmgr service
cfgmgr
The name of the configuration manager service. Currently ‘chef’ and ‘puppet’ are supported services.
cfgserver
The xCAT node name of the chef server or puppet master
roles
The roles associated with this node as recognized by the cfgmgr for the software that is to be installed and configured. These role names map to chef recipes or puppet manifest classes that should be used for this node. For example, chef OpenStack cookbooks have roles such as mysql-master,keystone, glance, nova-controller, nova-conductor, cinder-all.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
chain.5¶
chain - a table in the xCAT database.
chain Attributes: node, currstate, currchain, chain, ondiscover, comments, disable
Controls what operations are done (and it what order) when a node is discovered and deployed.
node
The node name or group name.
currstate
The current or next chain step to be executed on this node by xCAT-genesis. Set by xCAT during node discovery or as a result of nodeset.
currchain
The chain steps still left to do for this node. This attribute will be automatically adjusted by xCAT while xCAT-genesis is running on the node (either during node discovery or a special operation like firmware update). During node discovery, this attribute is initialized from the chain attribute and updated as the chain steps are executed.
chain
A comma-delimited chain of actions to be performed automatically when this node is discovered. (“Discovered” means a node booted, but xCAT and DHCP did not recognize the MAC of this node. In this situation, xCAT initiates the discovery process, the last step of which is to run the operations listed in this chain attribute, one by one.) Valid values: boot or reboot, install or netboot, runcmd=<cmd>, runimage=<URL>, shell, standby. (Default - same as no chain - it will do only the discovery.). Example, for BMC machines use: runcmd=bmcsetup,shell.
ondiscover
This attribute is currently not used by xCAT. The “nodediscover” operation is always done during node discovery.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
deps.5¶
deps - a table in the xCAT database.
deps Attributes: node, nodedep, msdelay, cmd, comments, disable
Describes dependencies some nodes have on others. This can be used, e.g., by rpower -d to power nodes on or off in the correct order.
node
The node name or group name.
nodedep
Comma-separated list of nodes or node groups it is dependent on.
msdelay
How long to wait between operating on the dependent nodes and the primary nodes.
cmd
Comma-seperated list of which operation this dependency applies to.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
discoverydata.5¶
discoverydata - a table in the xCAT database.
discoverydata Attributes: uuid, node, method, discoverytime, arch, cpucount, cputype, memory, mtm, serial, nicdriver, nicipv4, nichwaddr, nicpci, nicloc, niconboard, nicfirm, switchname, switchaddr, switchdesc, switchport, otherdata, comments, disable
Discovery data which sent from genesis.
uuid
The uuid of the node which send out the discovery request.
node
The node name which assigned to the discovered node.
method
The method which handled the discovery request. The method could be one of: switch, blade, profile, sequential.
discoverytime
The last time that xCAT received the discovery message.
arch
The architecture of the discovered node. e.g. x86_64.
cpucount
The number of cores multiply by threads core supported for the discovered node. e.g. 192.
cputype
The cpu type of the discovered node. e.g. Intel(R) Xeon(R) CPU E5-2690 0 @ 2.90GHz
memory
The memory size of the discovered node. e.g. 198460852
mtm
The machine type model of the discovered node. e.g. 786310X
serial
The serial number of the discovered node. e.g. 1052EFB
nicdriver
The driver of the nic. The value should be comma separated <nic name!driver name>. e.g. eth0!be2net,eth1!be2net
nicipv4
The ipv4 address of the nic. The value should be comma separated <nic name!ipv4 address>. e.g. eth0!10.0.0.212/8
nichwaddr
The hardware address of the nic. The should will be comma separated <nic name!hardware address>. e.g. eth0!34:40:B5:BE:DB:B0,eth1!34:40:B5:BE:DB:B4
nicpci
The pic device of the nic. The value should be comma separated <nic name!pci device>. e.g. eth0!0000:0c:00.0,eth1!0000:0c:00.1
nicloc
The location of the nic. The value should be comma separated <nic name!nic location>. e.g. eth0!Onboard Ethernet 1,eth1!Onboard Ethernet 2
niconboard
The onboard info of the nic. The value should be comma separated <nic name!onboard info>. e.g. eth0!1,eth1!2
nicfirm
The firmware description of the nic. The value should be comma separated <nic name!fimware description>. e.g. eth0!ServerEngines BE3 Controller,eth1!ServerEngines BE3 Controller
switchname
The switch name which the nic connected to. The value should be comma separated <nic name!switch name>. e.g. eth0!c909f06sw01
switchaddr
The address of the switch which the nic connected to. The value should be comma separated <nic name!switch address>. e.g. eth0!192.168.70.120
switchdesc
The description of the switch which the nic connected to. The value should be comma separated <nic name!switch description>. e.g. eth0!IBM Flex System Fabric EN4093 10Gb Scalable Switch, flash image: version 7.2.6, boot image: version 7.2.6
switchport
The port of the switch that the nic connected to. The value should be comma separated <nic name!switch port>. e.g. eth0!INTA2
otherdata
The left data which is not parsed to specific attributes (The complete message comes from genesis)
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
domain.5¶
domain - a table in the xCAT database.
domain Attributes: node, ou, authdomain, adminuser, adminpassword, type, comments, disable
Mapping of nodes to domain attributes
node
The node or group the entry applies to
ou
For an LDAP described machine account (i.e. Active Directory), the orginaztional unit to place the system. If not set, defaults to cn=Computers,dc=your,dc=domain
authdomain
If a node should participate in an AD domain or Kerberos realm distinct from domain indicated in site, this field can be used to specify that
adminuser
Allow a node specific indication of Administrative user. Most will want to just use passwd table to indicate this once rather than by node.
adminpassword
Allow a node specific indication of Administrative user password for the domain. Most will want to ignore this in favor of passwd table.
type
Type, if any, of authentication domain to manipulate. The only recognized value at the moment is activedirectory.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
eventlog.5¶
eventlog - a table in the xCAT database.
eventlog Attributes: recid, eventtime, eventtype, monitor, monnode, node, application, component, id, severity, message, rawdata, comments, disable
Stores the events occurred.
recid
The record id.
eventtime
The timestamp for the event.
eventtype
The type of the event.
monitor
The name of the monitor that monitors this event.
monnode
The node that monitors this event.
node
The node where the event occurred.
application
The application that reports the event.
component
The component where the event occurred.
id
The location or the resource name where the event occurred.
severity
The severity of the event. Valid values are: informational, warning, critical.
message
The full description of the event.
rawdata
The data that associated with the event.
comments
Any user-provided notes.
disable
Do not use. tabprune will not work if set to yes or 1
nodels(1), chtab(8), tabdump(8), tabedit(8)
firmware.5¶
firmware - a table in the xCAT database.
firmware Attributes: node, cfgfile, comments, disable
Maps node to firmware values to be used for setup at node discovery or later
node
The node id.
cfgfile
The file to use.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
hosts.5¶
hosts - a table in the xCAT database.
hosts Attributes: node, ip, hostnames, otherinterfaces, comments, disable
IP addresses and hostnames of nodes. This info is optional and is only used to populate /etc/hosts and DNS via makehosts and makedns. Using regular expressions in this table can be a quick way to populate /etc/hosts.
node
The node name or group name.
ip
The IP address of the node. This is only used in makehosts. The rest of xCAT uses system name resolution to resolve node names to IP addresses.
hostnames
Hostname aliases added to /etc/hosts for this node. Comma or blank separated list.
otherinterfaces
Other IP addresses to add for this node. Format: -<ext>:<ip>,<intfhostname>:<ip>,...
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
hwinv.5¶
hwinv - a table in the xCAT database.
hwinv Attributes: node, cputype, cpucount, memory, disksize, comments, disable
The hareware inventory for the node.
node
The node name or group name.
cputype
The cpu model name for the node.
cpucount
The number of cpus for the node.
memory
The size of the memory for the node in MB.
disksize
The size of the disks for the node in GB.
comments
Any user-provided notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
hypervisor.5¶
hypervisor - a table in the xCAT database.
hypervisor Attributes: node, type, mgr, interface, netmap, defaultnet, cluster, datacenter, preferdirect, comments, disable
Hypervisor parameters
node
The node or static group name
type
The plugin associated with hypervisor specific commands such as revacuate
mgr
The virtualization specific manager of this hypervisor when applicable
interface
The definition of interfaces for the hypervisor. The format is [networkname:interfacename:bootprotocol:IP:netmask:gateway] that split with | for each interface
netmap
Optional mapping of useful names to relevant physical ports. For example, 10ge=vmnic_16.0&vmnic_16.1,ge=vmnic1 would be requesting two virtual switches to be created, one called 10ge with vmnic_16.0 and vmnic_16.1 bonded, and another simply connected to vmnic1. Use of this allows abstracting guests from network differences amongst hypervisors
defaultnet
Optionally specify a default network entity for guests to join to if they do not specify.
cluster
Specify to the underlying virtualization infrastructure a cluster membership for the hypervisor.
datacenter
Optionally specify a datacenter for the hypervisor to exist in (only applicable to VMWare)
preferdirect
If a mgr is declared for a hypervisor, xCAT will default to using the mgr for all operations. If this is field is set to yes or 1, xCAT will prefer to directly communicate with the hypervisor if possible
comments
disable
nodels(1), chtab(8), tabdump(8), tabedit(8)
ipmi.5¶
ipmi - a table in the xCAT database.
ipmi Attributes: node, bmc, bmcport, taggedvlan, bmcid, username, password, comments, disable
Settings for nodes that are controlled by an on-board BMC via IPMI.
node
The node name or group name.
bmc
The hostname of the BMC adapater.
bmcport
In systems with selectable shared/dedicated ethernet ports, this parameter can be used to specify the preferred port. 0 means use the shared port, 1 means dedicated, blank is to not assign. The following special cases exist for IBM System x servers: For x3755 M3 systems, 0 means use the dedicated port, 1 means shared, blank is to not assign. For certain systems which have a mezzaine or ML2 adapter, there is a second value to include: For x3750 M4 (Model 8722): 0 2 1st 1Gbps interface for LOM 0 0 1st 10Gbps interface for LOM 0 3 2nd 1Gbps interface for LOM 0 1 2nd 10Gbps interface for LOM For x3750 M4 (Model 8752), x3850/3950 X6, dx360 M4, x3550 M4, and x3650 M4: 0 Shared (1st onboard interface) 1 Dedicated 2 0 First interface on ML2 or mezzanine adapter 2 1 Second interface on ML2 or mezzanine adapter 2 2 Third interface on ML2 or mezzanine adapter 2 3 Fourth interface on ML2 or mezzanine adapter
taggedvlan
Have bmcsetup place the BMC on the specified vlan tag on a shared netwirk interface. Some network devices may be incompatible with this option
bmcid
Unique identified data used by discovery processes to distinguish known BMCs from unrecognized BMCs
username
The BMC userid. If not specified, the key=ipmi row in the passwd table is used as the default.
password
The BMC password. If not specified, the key=ipmi row in the passwd table is used as the default.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
iscsi.5¶
iscsi - a table in the xCAT database.
iscsi Attributes: node, server, target, lun, iname, file, userid, passwd, kernel, kcmdline, initrd, comments, disable
Contains settings that control how to boot a node from an iSCSI target
node
The node name or group name.
server
The server containing the iscsi boot device for this node.
target
The iscsi disk used for the boot device for this node. Filled in by xCAT.
lun
LUN of boot device. Per RFC-4173, this is presumed to be 0 if unset. tgtd often requires this to be 1
iname
Initiator name. Currently unused.
file
The path on the server of the OS image the node should boot from.
userid
The userid of the iscsi server containing the boot device for this node.
passwd
The password for the iscsi server containing the boot device for this node.
kernel
The path of the linux kernel to boot from.
kcmdline
The kernel command line to use with iSCSI for this node.
initrd
The initial ramdisk to use when network booting this node.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
kit.5¶
kit - a table in the xCAT database.
kit Attributes: kitname, basename, description, version, release, ostype, isinternal, kitdeployparams, kitdir, comments, disable
This table stores all kits added to the xCAT cluster.
kitname
The unique generated kit name, when kit is added to the cluster.
basename
The kit base name
description
The Kit description.
version
The kit version
release
The kit release
ostype
The kit OS type. Linux or AIX.
isinternal
A flag to indicated if the Kit is internally used. When set to 1, the Kit is internal. If 0 or undefined, the kit is not internal.
kitdeployparams
The file containing the default deployment parameters for this Kit. These parameters are added to the OS Image definition.s list of deployment parameters when one or more Kit Components from this Kit are added to the OS Image.
kitdir
The path to Kit Installation directory on the Mgt Node.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
kitcomponent.5¶
kitcomponent - a table in the xCAT database.
kitcomponent Attributes: kitcompname, description, kitname, kitreponame, basename, version, release, serverroles, kitpkgdeps, prerequisite, driverpacks, kitcompdeps, postbootscripts, genimage_postinstall, exlist, comments, disable
This table stores all kit components added to the xCAT cluster.
kitcompname
The unique Kit Component name. It is auto-generated when the parent Kit is added to the cluster.
description
The Kit component description.
kitname
The Kit name which this Kit Component belongs to.
kitreponame
The Kit Package Repository name which this Kit Component belongs to.
basename
Kit Component basename.
version
Kit Component version.
release
Kit Component release.
serverroles
The types of servers that this Kit Component can install on. Valid types are: mgtnode, servicenode, compute
kitpkgdeps
Comma-separated list of packages that this kit component depends on.
prerequisite
Prerequisite for this kit component, the prerequisite includes ospkgdeps,preinstall,preupgrade,preuninstall scripts
driverpacks
Comma-separated List of driver package names. These must be full names like: pkg1-1.0-1.x86_64.rpm.
kitcompdeps
Comma-separated list of kit components that this kit component depends on.
postbootscripts
Comma-separated list of postbootscripts that will run during the node boot.
genimage_postinstall
Comma-separated list of postinstall scripts that will run during the genimage.
exlist
Exclude list file containing the files/directories to exclude when building a diskless image.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
kitrepo.5¶
kitrepo - a table in the xCAT database.
kitrepo Attributes: kitreponame, kitname, osbasename, osmajorversion, osminorversion, osarch, compat_osbasenames, kitrepodir, comments, disable
This table stores all kits added to the xCAT cluster.
kitreponame
The unique generated kit repo package name, when kit is added to the cluster.
kitname
The Kit name which this Kit Package Repository belongs to.
osbasename
The OS distro name which this repository is based on.
osmajorversion
The OS distro major version which this repository is based on.
osminorversion
The OS distro minor version which this repository is based on. If this attribute is not set, it means that this repo applies to all minor versions.
osarch
The OS distro arch which this repository is based on.
compat_osbasenames
List of compatible OS base names.
kitrepodir
The path to Kit Repository directory on the Mgt Node.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
kvm_masterdata.5¶
kvm_masterdata - a table in the xCAT database.
kvm_masterdata Attributes: name, xml, comments, disable
Persistant store for KVM plugin for masters
name
The name of the relevant master
xml
The XML description to be customized for clones of this master
comments
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
kvm_nodedata.5¶
kvm_nodedata - a table in the xCAT database.
kvm_nodedata Attributes: node, xml, comments, disable
Persistant store for KVM plugin, not intended for manual modification.
node
The node corresponding to the virtual machine
xml
The XML description generated by xCAT, fleshed out by libvirt, and stored for reuse
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
linuximage.5¶
linuximage - a table in the xCAT database.
linuximage Attributes: imagename, template, boottarget, addkcmdline, pkglist, pkgdir, otherpkglist, otherpkgdir, exlist, postinstall, rootimgdir, kerneldir, nodebootif, otherifce, netdrivers, kernelver, krpmver, permission, dump, crashkernelsize, partitionfile, driverupdatesrc, comments, disable
Information about a Linux operating system image that can be used to deploy cluster nodes.
imagename
The name of this xCAT OS image definition.
template
The fully qualified name of the template file that will be used to create the OS installer configuration file for stateful installations (e.g. kickstart for RedHat, autoyast for SLES).
boottarget
The name of the boottarget definition. When this attribute is set, xCAT will use the kernel, initrd and kernel params defined in the boottarget definition instead of the default.
addkcmdline
User specified arguments to be passed to the kernel. The user arguments are appended to xCAT.s default kernel arguments. For the kernel options need to be persistent after installation, specify them with prefix “R::”. This attribute is ignored if linuximage.boottarget is set.
pkglist
The fully qualified name of the file that stores the distro packages list that will be included in the image. Make sure that if the pkgs in the pkglist have dependency pkgs, the dependency pkgs should be found in one of the pkgdir
pkgdir
The name of the directory where the distro packages are stored. It could be set multiple paths.The multiple paths must be seperated by ”,”. The first path in the value of osimage.pkgdir must be the OS base pkg dir path, such as pkgdir=/install/rhels6.2/x86_64,/install/updates . In the os base pkg path, there are default repository data. And in the other pkg path(s), the users should make sure there are repository data. If not, use “createrepo” command to create them. For ubuntu, multiple mirrors can be specified in the pkgdir attribute, the mirrors must be prefixed by the protocol(http/ssh) and delimited with ”,” between each other.
otherpkglist
The fully qualified name of the file that stores non-distro package lists that will be included in the image. It could be set multiple paths.The multiple paths must be seperated by ”,”.
otherpkgdir
The base directory where the non-distro packages are stored. Only 1 local directory supported at present.
exlist
The fully qualified name of the file that stores the file names and directory names that will be excluded from the image during packimage command. It is used for diskless image only.
postinstall
The fully qualified name of the script file that will be run at the end of the genimage command. It could be set multiple paths.The multiple paths must be seperated by ”,”. It is used for diskless image only.
rootimgdir
The directory name where the image is stored. It is generally used for diskless image. it also can be used in sysclone environment to specify where the image captured from golden client is stored. in sysclone environment, rootimgdir is generally assigned to some default value by xcat, but you can specify your own store directory. just one thing need to be noticed, wherever you save the image, the name of last level directory must be the name of image. for example, if your image name is testimage and you want to save this image under home directoy, rootimgdir should be assigned to value /home/testimage/
kerneldir
The directory name where the 3rd-party kernel is stored. It is used for diskless image only.
nodebootif
The network interface the stateless/statelite node will boot over (e.g. eth0)
otherifce
Other network interfaces (e.g. eth1) in the image that should be configured via DHCP
netdrivers
The ethernet device drivers of the nodes which will use this linux image, at least the device driver for the nodes’ installnic should be included
kernelver
The version of linux kernel used in the linux image. If the kernel version is not set, the default kernel in rootimgdir will be used
krpmver
The rpm version of kernel packages (for SLES only). If it is not set, the default rpm version of kernel packages will be used.
permission
The mount permission of /.statelite directory is used, its default value is 755
dump
The NFS directory to hold the Linux kernel dump file (vmcore) when the node with this image crashes, its format is “nfs://<nfs_server_ip>/<kdump_path>”. If you want to use the node’s “xcatmaster” (its SN or MN), <nfs_server_ip> can be left blank. For example, “nfs:///<kdump_path>” means the NFS directory to hold the kernel dump file is on the node’s SN, or MN if there’s no SN.
crashkernelsize
the size that assigned to the kdump kernel. If the kernel size is not set, 256M will be the default value.
partitionfile
The path of the configuration file which will be used to partition the disk for the node. For stateful osimages,two types of files are supported: “<partition file absolute path>” which contains a partitioning definition that will be inserted directly into the generated autoinst configuration file and must be formatted for the corresponding OS installer (e.g. kickstart for RedHat, autoyast for SLES, pressed for Ubuntu). “s:<partitioning script absolute path>” which specifies a shell script that will be run from the OS installer configuration file %pre section; the script must write the correct partitioning definition into the file /tmp/partitionfile on the node which will be included into the configuration file during the install process. For statelite osimages, partitionfile should specify “<partition file absolute path>”; see the xCAT Statelite documentation for the xCAT defined format of this configuration file.For Ubuntu, besides “<partition file absolute path>” or “s:<partitioning script absolute path>”, the disk name(s) to partition must be specified in traditional, non-devfs format, delimited with space, it can be specified in 2 forms: “d:<the absolute path of the disk name file>” which contains the disk name(s) to partition and “s:d:<the absolute path of the disk script>” which runs in pressed/early_command and writes the disk names into the “/tmp/boot_disk” . To support other specific partition methods such as RAID or LVM in Ubuntu, some additional preseed values should be specified, these values can be specified with “c:<the absolute path of the additional pressed config file>” which contains the additional pressed entries in “d-i ...” form and “s:c:<the absolute path of the additional pressed config script>” which runs in pressed/early_command and set the preseed values with “debconf-set”. The multiple values should be delimited with comma ”,”
driverupdatesrc
The source of the drivers which need to be loaded during the boot. Two types of driver update source are supported: Driver update disk and Driver rpm package. The value for this attribute should be comma separated sources. Each source should be the format tab:full_path_of_srouce_file. The tab keyword can be: dud (for Driver update disk) and rpm (for driver rpm). If missing the tab, the rpm format is the default. e.g. dud:/install/dud/dd.img,rpm:/install/rpm/d.rpm
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
litefile.5¶
litefile - a table in the xCAT database.
litefile Attributes: image, file, options, comments, disable
The litefile table specifies the directories and files on the statelite nodes that should be readwrite, persistent, or readonly overlay. All other files in the statelite nodes come from the readonly statelite image.
image
The name of the image (as specified in the osimage table) that will use these options on this dir/file. You can also specify an image group name that is listed in the osimage.groups attribute of some osimages. ‘ALL’ means use this row for all images.
file
The full pathname of the file. e.g: /etc/hosts. If the path is a directory, then it should be terminated with a ‘/’.
options
Options for the file:
tmpfs - It is the default option if you leave the options column blank. It provides a file or directory for the node to use when booting, its permission will be the same as the original version on the server. In most cases, it is read-write; however, on the next statelite boot, the original version of the file or directory on the server will be used, it means it is non-persistent. This option can be performed on files and directories.. rw - Same as Above.Its name "rw" does NOT mean it always be read-write, even in most cases it is read-write. Please do not confuse it with the "rw" permission in the file system. persistent - It provides a mounted file or directory that is copied to the xCAT persistent location and then over-mounted on the local file or directory. Anything written to that file or directory is preserved. It means, if the file/directory does not exist at first, it will be copied to the persistent location. Next time the file/directory in the persistent location will be used. The file/directory will be persistent across reboots. Its permission will be the same as the original one in the statelite location. It requires the statelite table to be filled out with a spot for persistent statelite. This option can be performed on files and directories. con - The contents of the pathname are concatenated to the contents of the existing file. For this directive the searching in the litetree hierarchy does not stop when the first match is found. All files found in the hierarchy will be concatenated to the file when found. The permission of the file will be "-rw-r--r--", which means it is read-write for the root user, but readonly for the others. It is non-persistent, when the node reboots, all changes to the file will be lost. It can only be performed on files. Please do not use it for one directory. ro - The file/directory will be overmounted read-only on the local file/directory. It will be located in the directory hierarchy specified in the litetree table. Changes made to this file or directory on the server will be immediately seen in this file/directory on the node. This option requires that the file/directory to be mounted must be available in one of the entries in the litetree table. This option can be performed on files and directories. link - It provides one file/directory for the node to use when booting, it is copied from the server, and will be placed in tmpfs on the booted node. In the local file system of the booted node, it is one symbolic link to one file/directory in tmpfs. And the permission of the symbolic link is "lrwxrwxrwx", which is not the real permission of the file/directory on the node. So for some application sensitive to file permissions, it will be one issue to use "link" as its option, for example, "/root/.ssh/", which is used for SSH, should NOT use "link" as its option. It is non-persistent, when the node is rebooted, all changes to the file/directory will be lost. This option can be performed on files and directories. link,con - It works similar to the "con" option. All the files found in the litetree hierarchy will be concatenated to the file when found. The final file will be put to the tmpfs on the booted node. In the local file system of the booted node, it is one symbolic link to the file/directory in tmpfs. It is non-persistent, when the node is rebooted, all changes to the file will be lost. The option can only be performed on files. link,persistent - It provides a mounted file or directory that is copied to the xCAT persistent location and then over-mounted to the tmpfs on the booted node, and finally the symbolic link in the local file system will be linked to the over-mounted tmpfs file/directory on the booted node. The file/directory will be persistent across reboots. The permission of the file/directory where the symbolic link points to will be the same as the original one in the statelite location. It requires the statelite table to be filled out with a spot for persistent statelite. The option can be performed on files and directories. link,ro - The file is readonly, and will be placed in tmpfs on the booted node. In the local file system of the booted node, it is one symbolic link to the tmpfs. It is non-persistent, when the node is rebooted, all changes to the file/directory will be lost. This option requires that the file/directory to be mounted must be available in one of the entries in the litetree table. The option can be performed on files and directories.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
litetree.5¶
litetree - a table in the xCAT database.
litetree Attributes: priority, image, directory, mntopts, comments, disable
Directory hierarchy to traverse to get the initial contents of node files. The files that are specified in the litefile table are searched for in the directories specified in this table.
priority
This number controls what order the directories are searched. Directories are searched from smallest priority number to largest.
image
The name of the image (as specified in the osimage table) that will use this directory. You can also specify an image group name that is listed in the osimage.groups attribute of some osimages. ‘ALL’ means use this row for all images.
directory
The location (hostname:path) of a directory that contains files specified in the litefile table. Variables are allowed. E.g: $noderes.nfsserver://xcatmasternode/install/$node/#CMD=uname-r#/
mntopts
A comma-separated list of options to use when mounting the litetree directory. (Ex. ‘soft’) The default is to do a ‘hard’ mount.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
mac.5¶
mac - a table in the xCAT database.
mac Attributes: node, interface, mac, comments, disable
The MAC address of the node’s install adapter. Normally this table is populated by getmacs or node discovery, but you can also add entries to it manually.
node
The node name or group name.
interface
The adapter interface name that will be used to install and manage the node. E.g. eth0 (for linux) or en0 (for AIX).)
mac
The mac address or addresses for which xCAT will manage static bindings for this node. This may be simply a mac address, which would be bound to the node name (such as “01:02:03:04:05:0E”). This may also be a “|” delimited string of “mac address!hostname” format (such as “01:02:03:04:05:0E!node5|01:02:03:05:0F!node6-eth1”).
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
mic.5¶
mic - a table in the xCAT database.
mic Attributes: node, host, id, nodetype, bridge, onboot, vlog, powermgt, comments, disable
The host, slot id and configuraton of the mic (Many Integrated Core).
node
The node name or group name.
host
The host node which the mic card installed on.
id
The device id of the mic node.
nodetype
The hardware type of the mic node. Generally, it is mic.
bridge
The virtual bridge on the host node which the mic connected to.
onboot
Set mic to autoboot when mpss start. Valid values: yes|no. Default is yes.
vlog
Set the Verbose Log to console. Valid values: yes|no. Default is no.
powermgt
Set the Power Management for mic node. This attribute is used to set the power management state that mic may get into when it is idle. Four states can be set: cpufreq, corec6, pc3 and pc6. The valid value for powermgt attribute should be [cpufreq=<on|off>]![corec6=<on|off>]![pc3=<on|off>]![pc6=<on|off>]. e.g. cpufreq=on!corec6=off!pc3=on!pc6=off. Refer to the doc of mic to get more information for power management.
comments
Any user-provided notes.
disable
Do not use. tabprune will not work if set to yes or 1
nodels(1), chtab(8), tabdump(8), tabedit(8)
monitoring.5¶
monitoring - a table in the xCAT database.
monitoring Attributes: name, nodestatmon, comments, disable
Controls what external monitoring tools xCAT sets up and uses. Entries should be added and removed from this table using the provided xCAT commands monstart and monstop.
name
The name of the mornitoring plug-in module. The plug-in must be put in /lib/perl/xCAT_monitoring/. See the man page for monstart for details.
nodestatmon
Specifies if the monitoring plug-in is used to feed the node status to the xCAT cluster. Any one of the following values indicates “yes”: y, Y, yes, Yes, YES, 1. Any other value or blank (default), indicates “no”.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
monsetting.5¶
monsetting - a table in the xCAT database.
monsetting Attributes: name, key, value, comments, disable
Specifies the monitoring plug-in specific settings. These settings will be used by the monitoring plug-in to customize the behavior such as event filter, sample interval, responses etc. Entries should be added, removed or modified by chtab command. Entries can also be added or modified by the monstart command when a monitoring plug-in is brought up.
name
The name of the mornitoring plug-in module. The plug-in must be put in /lib/perl/xCAT_monitoring/. See the man page for monstart for details.
key
Specifies the name of the attribute. The valid values are specified by each monitoring plug-in. Use “monls name -d” to get a list of valid keys.
value
Specifies the value of the attribute.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
mp.5¶
mp - a table in the xCAT database.
mp Attributes: node, mpa, id, nodetype, comments, disable
Contains the hardware control info specific to blades. This table also refers to the mpa table, which contains info about each Management Module.
node
The blade node name or group name.
mpa
The managment module used to control this blade.
id
The slot number of this blade in the BladeCenter chassis.
nodetype
The hardware type for mp node. Valid values: mm,cmm, blade.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
mpa.5¶
mpa - a table in the xCAT database.
mpa Attributes: mpa, username, password, displayname, slots, urlpath, comments, disable
Contains info about each Management Module and how to access it.
mpa
Hostname of the management module.
username
Userid to use to access the management module.
password
Password to use to access the management module. If not specified, the key=blade row in the passwd table is used as the default.
displayname
Alternative name for BladeCenter chassis. Only used by PCM.
slots
- The number of available slots in the chassis. For PCM, this attribute is used to store the number of slots in the following format: <slot rows>,<slot columns>,<slot orientation> Where:
- <slot rows> = number of rows of slots in chassis <slot columns> = number of columns of slots in chassis <slot orientation> = set to 0 if slots are vertical, and set to 1 if slots of horizontal
urlpath
URL path for the Chassis web interface. The full URL is built as follows: <hostname>/<urlpath>
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
networks.5¶
networks - a table in the xCAT database.
networks Attributes: netname, net, mask, mgtifname, gateway, dhcpserver, tftpserver, nameservers, ntpservers, logservers, dynamicrange, staticrange, staticrangeincrement, nodehostname, ddnsdomain, vlanid, domain, comments, disable
Describes the networks in the cluster and info necessary to set up nodes on that network.
netname
Name used to identify this network definition.
net
The network address.
mask
The network mask.
mgtifname
The interface name of the management/service node facing this network. !remote!<nicname> indicates a non-local network on a specific nic for relay DHCP.
gateway
The network gateway. It can be set to an ip address or the keyword <xcatmaster>, the keyword <xcatmaster> indicates the cluster-facing ip address configured on this management node or service node. Leaving this field blank means that there is no gateway for this network.
dhcpserver
The DHCP server that is servicing this network. Required to be explicitly set for pooled service node operation.
tftpserver
The TFTP server that is servicing this network. If not set, the DHCP server is assumed.
nameservers
A comma delimited list of DNS servers that each node in this network should use. This value will end up in the nameserver settings of the /etc/resolv.conf on each node in this network. If this attribute value is set to the IP address of an xCAT node, make sure DNS is running on it. In a hierarchical cluster, you can also set this attribute to “<xcatmaster>” to mean the DNS server for each node in this network should be the node that is managing it (either its service node or the management node). Used in creating the DHCP network definition, and DNS configuration.
ntpservers
The ntp servers for this network. Used in creating the DHCP network definition. Assumed to be the DHCP server if not set.
logservers
The log servers for this network. Used in creating the DHCP network definition. Assumed to be the DHCP server if not set.
dynamicrange
The IP address range used by DHCP to assign dynamic IP addresses for requests on this network. This should not overlap with entities expected to be configured with static host declarations, i.e. anything ever expected to be a node with an address registered in the mac table.
staticrange
The IP address range used to dynamically assign static IPs to newly discovered nodes. This should not overlap with the dynamicrange nor overlap with entities that were manually assigned static IPs. The format for the attribute value is: <startip>-<endip>.
staticrangeincrement
nodehostname
A regular expression used to specify node name to network-specific hostname. i.e. “/z/-secondary/” would mean that the hostname of “n1” would be n1-secondary on this network. By default, the nodename is assumed to equal the hostname, followed by nodename-interfacename.
ddnsdomain
A domain to be combined with nodename to construct FQDN for DDNS updates induced by DHCP. This is not passed down to the client as “domain”
vlanid
The vlan ID if this network is within a vlan.
domain
The DNS domain name (ex. cluster.com).
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
nics.5¶
nics - a table in the xCAT database.
nics Attributes: node, nicips, nichostnamesuffixes, nichostnameprefixes, nictypes, niccustomscripts, nicnetworks, nicaliases, nicextraparams, comments, disable
Stores NIC details.
node
The node or group name.
nicips
- Comma-separated list of IP addresses per NIC. To specify one ip address per NIC:
- To specify multiple ip addresses per NIC:
- <nic1>!<ip1>|<ip2>,<nic2>!<ip1>|<ip2>,..., for example, eth0!10.0.0.100|fd55::214:5eff:fe15:849b,ib0!11.0.0.100|2001::214:5eff:fe15:849a. The xCAT object definition commands support to use nicips.<nicname> as the sub attributes.
Note: The primary IP address must also be stored in the hosts.ip attribute. The nichostnamesuffixes should specify one hostname suffix for each ip address.
nichostnamesuffixes
- Comma-separated list of hostname suffixes per NIC.
- If only one ip address is associated with each NIC:
- <nic1>!<ext1>,<nic2>!<ext2>,..., for example, eth0!-eth0,ib0!-ib0
- If multiple ip addresses are associcated with each NIC:
- <nic1>!<ext1>|<ext2>,<nic2>!<ext1>|<ext2>,..., for example, eth0!-eth0|-eth0-ipv6,ib0!-ib0|-ib0-ipv6.
The xCAT object definition commands support to use nichostnamesuffixes.<nicname> as the sub attributes. Note: According to DNS rules a hostname must be a text string up to 24 characters drawn from the alphabet (A-Z), digits (0-9), minus sign (-),and period (.). When you are specifying “nichostnamesuffixes” or “nicaliases” make sure the resulting hostnames will conform to this naming convention
nichostnameprefixes
- Comma-separated list of hostname prefixes per NIC.
- If only one ip address is associated with each NIC:
- <nic1>!<ext1>,<nic2>!<ext2>,..., for example, eth0!eth0-,ib0!ib-
- If multiple ip addresses are associcated with each NIC:
- <nic1>!<ext1>|<ext2>,<nic2>!<ext1>|<ext2>,..., for example, eth0!eth0-|eth0-ipv6i-,ib0!ib-|ib-ipv6-.
The xCAT object definition commands support to use nichostnameprefixes.<nicname> as the sub attributes. Note: According to DNS rules a hostname must be a text string up to 24 characters drawn from the alphabet (A-Z), digits (0-9), minus sign (-),and period (.). When you are specifying “nichostnameprefixes” or “nicaliases” make sure the resulting hostnames will conform to this naming convention
nictypes
Comma-separated list of NIC types per NIC. <nic1>!<type1>,<nic2>!<type2>, e.g. eth0!Ethernet,ib0!Infiniband. The xCAT object definition commands support to use nictypes.<nicname> as the sub attributes.
niccustomscripts
Comma-separated list of custom scripts per NIC. <nic1>!<script1>,<nic2>!<script2>, e.g. eth0!configeth eth0, ib0!configib ib0. The xCAT object definition commands support to use niccustomscripts.<nicname> as the sub attribute .
nicnetworks
- Comma-separated list of networks connected to each NIC.
- If only one ip address is associated with each NIC:
- <nic1>!<network1>,<nic2>!<network2>, for example, eth0!10_0_0_0-255_255_0_0, ib0!11_0_0_0-255_255_0_0
- If multiple ip addresses are associated with each NIC:
- <nic1>!<network1>|<network2>,<nic2>!<network1>|<network2>, for example, eth0!10_0_0_0-255_255_0_0|fd55:faaf:e1ab:336::/64,ib0!11_0_0_0-255_255_0_0|2001:db8:1:0::/64. The xCAT object definition commands support to use nicnetworks.<nicname> as the sub attributes.
nicaliases
- Comma-separated list of hostname aliases for each NIC.
- Format: eth0!<alias list>,eth1!<alias1 list>|<alias2 list>
- For multiple aliases per nic use a space-separated list.
For example: eth0!moe larry curly,eth1!tom|jerry
nicextraparams
- Comma-separated list of extra parameters that will be used for each NIC configuration.
- If only one ip address is associated with each NIC:
- <nic1>!<param1=value1 param2=value2>,<nic2>!<param3=value3>, for example, eth0!MTU=1500,ib0!MTU=65520 CONNECTED_MODE=yes.
- If multiple ip addresses are associated with each NIC:
- <nic1>!<param1=value1 param2=value2>|<param3=value3>,<nic2>!<param4=value4 param5=value5>|<param6=value6>, for example, eth0!MTU=1500|MTU=1460,ib0!MTU=65520 CONNECTED_MODE=yes.
The xCAT object definition commands support to use nicextraparams.<nicname> as the sub attributes.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
nimimage.5¶
nimimage - a table in the xCAT database.
nimimage Attributes: imagename, nimtype, lpp_source, spot, root, dump, paging, resolv_conf, tmp, home, shared_home, res_group, nimmethod, script, bosinst_data, installp_bundle, mksysb, fb_script, shared_root, otherpkgs, image_data, configdump, comments, disable
All the info that specifies a particular AIX operating system image that can be used to deploy AIX nodes.
imagename
User provided name of this xCAT OS image definition.
nimtype
The NIM client type- standalone, diskless, or dataless.
lpp_source
The name of the NIM lpp_source resource.
spot
The name of the NIM SPOT resource.
root
The name of the NIM root resource.
dump
The name of the NIM dump resource.
paging
The name of the NIM paging resource.
resolv_conf
The name of the NIM resolv_conf resource.
tmp
The name of the NIM tmp resource.
home
The name of the NIM home resource.
shared_home
The name of the NIM shared_home resource.
res_group
The name of a NIM resource group.
nimmethod
The NIM install method to use, (ex. rte, mksysb).
script
The name of a NIM script resource.
bosinst_data
The name of a NIM bosinst_data resource.
installp_bundle
One or more comma separated NIM installp_bundle resources.
mksysb
The name of a NIM mksysb resource.
fb_script
The name of a NIM fb_script resource.
shared_root
A shared_root resource represents a directory that can be used as a / (root) directory by one or more diskless clients.
otherpkgs
One or more comma separated installp or rpm packages. The rpm packages must have a prefix of ‘R:’, (ex. R:foo.rpm)
image_data
The name of a NIM image_data resource.
configdump
Specifies the type of system dump to be collected. The values are selective, full, and none. The default is selective.
comments
Any user-provided notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
nodegroup.5¶
nodegroup - a table in the xCAT database.
nodegroup Attributes: groupname, grouptype, members, membergroups, wherevals, comments, disable
Contains group definitions, whose membership is dynamic depending on characteristics of the node.
groupname
Name of the group.
grouptype
The only current valid value is dynamic. We will be looking at having the object def commands working with static group definitions in the nodelist table.
members
The value of the attribute is not used, but the attribute is necessary as a place holder for the object def commands. (The membership for static groups is stored in the nodelist table.)
membergroups
This attribute stores a comma-separated list of nodegroups that this nodegroup refers to. This attribute is only used by PCM.
wherevals
A list of “attr*val” pairs that can be used to determine the members of a dynamic group, the delimiter is ”::” and the operator * can be ==, =~, != or !~.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
nodehm.5¶
nodehm - a table in the xCAT database.
nodehm Attributes: node, power, mgt, cons, termserver, termport, conserver, serialport, serialspeed, serialflow, getmac, cmdmapping, consoleondemand, comments, disable
Settings that control how each node’s hardware is managed. Typically, an additional table that is specific to the hardware type of the node contains additional info. E.g. the ipmi, mp, and ppc tables.
node
The node name or group name.
power
The method to use to control the power of the node. If not set, the mgt attribute will be used. Valid values: ipmi, blade, hmc, ivm, fsp, kvm, esx, rhevm. If “ipmi”, xCAT will search for this node in the ipmi table for more info. If “blade”, xCAT will search for this node in the mp table. If “hmc”, “ivm”, or “fsp”, xCAT will search for this node in the ppc table.
mgt
The method to use to do general hardware management of the node. This attribute is used as the default if power or getmac is not set. Valid values: ipmi, blade, hmc, ivm, fsp, bpa, kvm, esx, rhevm. See the power attribute for more details.
cons
The console method. If nodehm.serialport is set, this will default to the nodehm.mgt setting, otherwise it defaults to unused. Valid values: cyclades, mrv, or the values valid for the mgt attribute.
termserver
The hostname of the terminal server.
termport
The port number on the terminal server that this node is connected to.
conserver
The hostname of the machine where the conserver daemon is running. If not set, the default is the xCAT management node.
serialport
The serial port for this node, in the linux numbering style (0=COM1/ttyS0, 1=COM2/ttyS1). For SOL on IBM blades, this is typically 1. For rackmount IBM servers, this is typically 0.
serialspeed
The speed of the serial port for this node. For SOL this is typically 19200.
serialflow
The flow control value of the serial port for this node. For SOL this is typically ‘hard’.
getmac
The method to use to get MAC address of the node with the getmac command. If not set, the mgt attribute will be used. Valid values: same as values for mgmt attribute.
cmdmapping
The fully qualified name of the file that stores the mapping between PCM hardware management commands and xCAT/third-party hardware management commands for a particular type of hardware device. Only used by PCM.
consoleondemand
This overrides the value from site.consoleondemand; (0=no, 1=yes). Default is the result from site.consoleondemand.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
nodelist.5¶
nodelist - a table in the xCAT database.
nodelist Attributes: node, groups, status, statustime, appstatus, appstatustime, primarysn, hidden, updatestatus, updatestatustime, zonename, comments, disable
The list of all the nodes in the cluster, including each node’s current status and what groups it is in.
node
The hostname of a node in the cluster.
groups
A comma-delimited list of groups this node is a member of. Group names are arbitrary, except all nodes should be part of the ‘all’ group. Internal group names are designated by using __<groupname>. For example, __Unmanaged, could be the internal name for a group of nodes that is not managed by xCAT. Admins should avoid using the __ characters when defining their groups.
status
The current status of this node. This attribute will be set by xCAT software. Valid values: defined, booting, netbooting, booted, discovering, configuring, installing, alive, standingby, powering-off, unreachable. If blank, defined is assumed. The possible status change sequenses are: For installaton: defined->[discovering]->[configuring]->[standingby]->installing->booting->booted->[alive], For diskless deployment: defined->[discovering]->[configuring]->[standingby]->netbooting->booted->[alive], For booting: [alive/unreachable]->booting->[alive], For powering off: [alive]->powering-off->[unreachable], For monitoring: alive->unreachable. Discovering and configuring are for x Series dicovery process. Alive and unreachable are set only when there is a monitoring plug-in start monitor the node status for xCAT. Please note that the status values will not reflect the real node status if you change the state of the node from outside of xCAT (i.e. power off the node using HMC GUI).
statustime
The data and time when the status was updated.
appstatus
A comma-delimited list of application status. For example: ‘sshd=up,ftp=down,ll=down’
appstatustime
The date and time when appstatus was updated.
primarysn
Not used currently. The primary servicenode, used by this node.
hidden
Used to hide fsp and bpa definitions, 1 means not show them when running lsdef and nodels
updatestatus
The current node update status. Valid states are synced, out-of-sync,syncing,failed.
updatestatustime
The date and time when the updatestatus was updated.
zonename
The name of the zone to which the node is currently assigned. If undefined, then it is not assigned to any zone.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
nodepos.5¶
nodepos - a table in the xCAT database.
nodepos Attributes: node, rack, u, chassis, slot, room, height, comments, disable
Contains info about the physical location of each node. Currently, this info is not used by xCAT, and therefore can be in whatevery format you want. It will likely be used in xCAT in the future.
node
The node name or group name.
rack
The frame the node is in.
u
The vertical position of the node in the frame
chassis
The BladeCenter chassis the blade is in.
slot
The slot number of the blade in the chassis. For PCM, a comma-separated list of slot numbers is stored
room
The room where the node is located.
height
The server height in U(s).
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
noderes.5¶
noderes - a table in the xCAT database.
noderes Attributes: node, servicenode, netboot, tftpserver, tftpdir, nfsserver, monserver, nfsdir, installnic, primarynic, discoverynics, cmdinterface, xcatmaster, current_osimage, next_osimage, nimserver, routenames, nameservers, proxydhcp, comments, disable
Resources and settings to use when installing nodes.
node
The node name or group name.
servicenode
A comma separated list of node names (as known by the management node) that provides most services for this node. The first service node on the list that is accessible will be used. The 2nd node on the list is generally considered to be the backup service node for this node when running commands like snmove.
netboot
The type of network booting to use for this node. Valid values: pxe or xnba for x86* architecture, yaboot for POWER architecture, grub2-tftp and grub2-http for RHEL7 on Power and all the os deployment on Power LE. Notice: yaboot is not supported from rhels7 on Power,use grub2-tftp or grub2-http instead, the difference between the 2 is the file transfer protocol(i.e, http or tftp)
tftpserver
The TFTP server for this node (as known by this node). If not set, it defaults to networks.tftpserver.
tftpdir
The directory that roots this nodes contents from a tftp and related perspective. Used for NAS offload by using different mountpoints.
nfsserver
The NFS or HTTP server for this node (as known by this node).
monserver
The monitoring aggregation point for this node. The format is “x,y” where x is the ip address as known by the management node and y is the ip address as known by the node.
nfsdir
The path that should be mounted from the NFS server.
installnic
The network adapter on the node that will be used for OS deployment, the installnic can be set to the network adapter name or the mac address or the keyword “mac” which means that the network interface specified by the mac address in the mac table will be used. If not set, primarynic will be used. If primarynic is not set too, the keyword “mac” will be used as default.
primarynic
This attribute will be deprecated. All the used network interface will be determined by installnic. The network adapter on the node that will be used for xCAT management, the primarynic can be set to the network adapter name or the mac address or the keyword “mac” which means that the network interface specified by the mac address in the mac table will be used. Default is eth0.
discoverynics
If specified, force discovery to occur on specific network adapters only, regardless of detected connectivity. Syntax can be simply “eth2,eth3” to restrict discovery to whatever happens to come up as eth2 and eth3, or by driver name such as “bnx2:0,bnx2:1” to specify the first two adapters managed by the bnx2 driver
cmdinterface
Not currently used.
xcatmaster
The hostname of the xCAT service node (as known by this node). This acts as the default value for nfsserver and tftpserver, if they are not set. If xcatmaster is not set, the node will use whoever responds to its boot request as its master. For the directed bootp case for POWER, it will use the management node if xcatmaster is not set.
current_osimage
Not currently used. The name of the osimage data object that represents the OS image currently deployed on this node.
next_osimage
Not currently used. The name of the osimage data object that represents the OS image that will be installed on the node the next time it is deployed.
nimserver
Not used for now. The NIM server for this node (as known by this node).
routenames
A comma separated list of route names that refer to rows in the routes table. These are the routes that should be defined on this node when it is deployed.
nameservers
An optional node/group specific override for name server list. Most people want to stick to site or network defined nameserver configuration.
proxydhcp
To specify whether the node supports proxydhcp protocol. Valid values: yes or 1, no or 0. Default value is yes.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
nodetype.5¶
nodetype - a table in the xCAT database.
nodetype Attributes: node, os, arch, profile, provmethod, supportedarchs, nodetype, comments, disable
A few hardware and software characteristics of the nodes.
node
The node name or group name.
os
The operating system deployed on this node. Valid values: AIX, rhels*,rhelc*, rhas*,centos*,SL*, fedora*, sles* (where * is the version #). As a special case, if this is set to “boottarget”, then it will use the initrd/kernel/parameters specified in the row in the boottarget table in which boottarget.bprofile equals nodetype.profile.
arch
The hardware architecture of this node. Valid values: x86_64, ppc64, x86, ia64.
profile
The string to use to locate a kickstart or autoyast template to use for OS deployment of this node. If the provmethod attribute is set to an osimage name, that takes precedence, and profile need not be defined. Otherwise, the os, profile, and arch are used to search for the files in /install/custom first, and then in /opt/xcat/share/xcat.
provmethod
The provisioning method for node deployment. The valid values are install, netboot, statelite or an os image name from the osimage table. If an image name is specified, the osimage definition stored in the osimage table and the linuximage table (for Linux) or nimimage table (for AIX) are used to locate the files for templates, pkglists, syncfiles, etc. On Linux, if install, netboot or statelite is specified, the os, profile, and arch are used to search for the files in /install/custom first, and then in /opt/xcat/share/xcat.
supportedarchs
Comma delimited list of architectures this node can execute.
nodetype
A comma-delimited list of characteristics of this node. Valid values: ppc, blade, vm (virtual machine), osi (OS image), mm, mn, rsa, switch.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
notification.5¶
notification - a table in the xCAT database.
notification Attributes: filename, tables, tableops, comments, disable
Contains registrations to be notified when a table in the xCAT database changes. Users can add entries to have additional software notified of changes. Add and remove entries using the provided xCAT commands regnotif and unregnotif.
filename
The path name of a file that implements the callback routine when the monitored table changes. Can be a perl module or a command. See the regnotif man page for details.
tables
Comma-separated list of xCAT database tables to monitor.
tableops
Specifies the table operation to monitor for. Valid values: “d” (rows deleted), “a” (rows added), “u” (rows updated).
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
osdistro.5¶
osdistro - a table in the xCAT database.
osdistro Attributes: osdistroname, basename, majorversion, minorversion, arch, type, dirpaths, comments, disable
Information about all the OS distros in the xCAT cluster
osdistroname
Unique name (e.g. rhels6.2-x86_64)
basename
The OS base name (e.g. rhels)
majorversion
The OS distro major version.(e.g. 6)
minorversion
The OS distro minor version. (e.g. 2)
arch
The OS distro arch (e.g. x86_64)
type
Linux or AIX
dirpaths
Directory paths where OS distro is store. There could be multiple paths if OS distro has more than one ISO image. (e.g. /install/rhels6.2/x86_64,...)
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
osdistroupdate.5¶
osdistroupdate - a table in the xCAT database.
osdistroupdate Attributes: osupdatename, osdistroname, dirpath, downloadtime, comments, disable
Information about the OS distro updates in the xCAT cluster
osupdatename
Name of OS update. (e.g. rhn-update1)
osdistroname
The OS distro name to update. (e.g. rhels)
dirpath
Path to where OS distro update is stored. (e.g. /install/osdistroupdates/rhels6.2-x86_64-20120716-update)
downloadtime
The timestamp when OS distro update was downloaded..
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
osimage.5¶
osimage - a table in the xCAT database.
osimage Attributes: imagename, groups, profile, imagetype, description, provmethod, rootfstype, osdistroname, osupdatename, cfmdir, osname, osvers, osarch, synclists, postscripts, postbootscripts, serverrole, isdeletable, kitcomponents, comments, disable
Basic information about an operating system image that can be used to deploy cluster nodes.
imagename
The name of this xCAT OS image definition.
groups
A comma-delimited list of image groups of which this image is a member. Image groups can be used in the litefile and litetree table instead of a single image name. Group names are arbitrary.
profile
The node usage category. For example compute, service.
imagetype
The type of operating system image this definition represents (linux,AIX).
description
OS Image Description
provmethod
The provisioning method for node deployment. The valid values are install, netboot,statelite,boottarget,dualboot,sysclone. If boottarget is set, you must set linuximage.boottarget to the name of the boottarget definition. It is not used by AIX.
rootfstype
The filesystem type for the rootfs is used when the provmethod is statelite. The valid values are nfs or ramdisk. The default value is nfs
osdistroname
The name of the OS distro definition. This attribute can be used to specify which OS distro to use, instead of using the osname,osvers,and osarch attributes. For *kit commands, the attribute will be used to read the osdistro table for the osname, osvers, and osarch attributes. If defined, the osname, osvers, and osarch attributes defined in the osimage table will be ignored.
osupdatename
A comma-separated list of OS distro updates to apply to this osimage.
cfmdir
CFM directory name for PCM. Set to /install/osimages/<osimage name>/cfmdir by PCM.
osname
Operating system name- AIX or Linux.
osvers
The Linux operating system deployed on this node. Valid values: rhels*,rhelc*, rhas*,centos*,SL*, fedora*, sles* (where * is the version #).
osarch
The hardware architecture of this node. Valid values: x86_64, ppc64, x86, ia64.
synclists
The fully qualified name of a file containing a list of files to synchronize on the nodes. Can be a comma separated list of multiple synclist files. The synclist generated by PCM named /install/osimages/<imagename>/synclist.cfm is reserved for use only by PCM and should not be edited by the admin.
postscripts
Comma separated list of scripts that should be run on this image after diskfull installation or diskless boot. For installation of RedHat, CentOS, Fedora, the scripts will be run before the reboot. For installation of SLES, the scripts will be run after the reboot but before the init.d process. For diskless deployment, the scripts will be run at the init.d time, and xCAT will automatically add the list of scripts from the postbootscripts attribute to run after postscripts list. For installation of AIX, the scripts will run after the reboot and acts the same as the postbootscripts attribute. For AIX, use the postbootscripts attribute. See the site table runbootscripts attribute. Support will be added in the future for the postscripts attribute to run the scripts before the reboot in AIX.
postbootscripts
Comma separated list of scripts that should be run on this after diskfull installation or diskless boot. On AIX these scripts are run during the processing of /etc/inittab. On Linux they are run at the init.d time. xCAT automatically adds the scripts in the xcatdefaults.postbootscripts attribute to run first in the list. See the site table runbootscripts attribute.
serverrole
The role of the server created by this osimage. Default roles: mgtnode, servicenode, compute, login, storage, utility.
isdeletable
A flag to indicate whether this image profile can be deleted. This attribute is only used by PCM.
kitcomponents
List of Kit Component IDs assigned to this OS Image definition.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
passwd.5¶
passwd - a table in the xCAT database.
passwd Attributes: key, username, password, cryptmethod, authdomain, comments, disable
Contains default userids and passwords for xCAT to access cluster components. In most cases, xCAT will also actually set the userid/password in the relevant component when it is being configured or installed. Userids/passwords for specific cluster components can be overidden in other tables, e.g. mpa, ipmi, ppchcp, etc.
key
The type of component this user/pw is for. Valid values: blade (management module), ipmi (BMC), system (nodes), omapi (DHCP), hmc, ivm, cec, frame, switch.
username
The default userid for this type of component
password
The default password for this type of component
cryptmethod
Indicates the method that was used to encrypt the password attribute. On AIX systems, if a value is provided for this attribute it indicates that the password attribute is encrypted. If the cryptmethod value is not set it indicates the password is a simple string value. On Linux systems, the cryptmethod is not supported however the code attempts to auto-discover MD5 encrypted passwords.
authdomain
The domain in which this entry has meaning, e.g. specifying different domain administrators per active directory domain
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
performance.5¶
performance - a table in the xCAT database.
performance Attributes: timestamp, node, attrname, attrvalue, comments, disable
Describes the system performance every interval unit of time.
timestamp
The time at which the metric was captured.
node
The node name.
attrname
The metric name.
attrvalue
The metric value.
comments
Any user-provided notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
policy.5¶
policy - a table in the xCAT database.
policy Attributes: priority, name, host, commands, noderange, parameters, time, rule, comments, disable
The policy table in the xCAT database controls who has authority to run specific xCAT operations. It is basically the Access Control List (ACL) for xCAT. It is sorted on the priority field before evaluating.
priority
The priority value for this rule. This value is used to identify this policy data object (i.e. this rule) The table is sorted on this field with the lower the number the higher the priority. For example 1.0 is higher priority than 4.1 is higher than 4.9.
name
The username that is allowed to perform the commands specified by this rule. Default is “*” (all users).
host
The host from which users may issue the commands specified by this rule. Default is “*” (all hosts). Only all or one host is supported
commands
The list of commands that this rule applies to. Default is “*” (all commands).
noderange
The Noderange that this rule applies to. Default is “*” (all nodes). Not supported with the *def commands.
parameters
A regular expression that matches the command parameters (everything except the noderange) that this rule applies to. Default is “*” (all parameters). Not supported with the *def commands.
time
Time ranges that this command may be executed in. This is not supported.
rule
Specifies how this rule should be applied. Valid values are: allow, accept, trusted. Allow or accept will allow the user to run the commands. Any other value will deny the user access to the commands. Trusted means that once this client has been authenticated via the certificate, all other information that is sent (e.g. the username) is believed without question. This authorization should only be given to the xcatd on the management node at this time.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
postscripts.5¶
postscripts - a table in the xCAT database.
postscripts Attributes: node, postscripts, postbootscripts, comments, disable
The scripts that should be run on each node after installation or diskless boot.
node
The node name or group name.
postscripts
Comma separated list of scripts that should be run on this node after diskfull installation or diskless boot. Each script can take zero or more parameters. For example: “script1 p1 p2,script2,...”. xCAT automatically adds the postscripts from the xcatdefaults.postscripts attribute of the table to run first on the nodes after install or diskless boot. For installation of RedHat, CentOS, Fedora, the scripts will be run before the reboot. For installation of SLES, the scripts will be run after the reboot but before the init.d process. For diskless deployment, the scripts will be run at the init.d time, and xCAT will automatically add the list of scripts from the postbootscripts attribute to run after postscripts list. For installation of AIX, the scripts will run after the reboot and acts the same as the postbootscripts attribute. For AIX, use the postbootscripts attribute. Support will be added in the future for the postscripts attribute to run the scripts before the reboot in AIX.
postbootscripts
Comma separated list of scripts that should be run on this node after diskfull installation or diskless boot. Each script can take zero or more parameters. For example: “script1 p1 p2,script2,...”. On AIX these scripts are run during the processing of /etc/inittab. On Linux they are run at the init.d time. xCAT automatically adds the scripts in the xcatdefaults.postbootscripts attribute to run first in the list.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
ppc.5¶
ppc - a table in the xCAT database.
ppc Attributes: node, hcp, id, pprofile, parent, nodetype, supernode, sfp, comments, disable
List of system p hardware: HMCs, IVMs, FSPs, BPCs, CECs, Frames.
node
The node name or group name.
hcp
The hardware control point for this node (HMC, IVM, Frame or CEC). Do not need to set for BPAs and FSPs.
id
For LPARs: the LPAR numeric id; for CECs: the cage number; for Frames: the frame number.
pprofile
The LPAR profile that will be used the next time the LPAR is powered on with rpower. For DFM, the pprofile attribute should be set to blank
parent
For LPARs: the CEC; for FSPs: the CEC; for CEC: the frame (if one exists); for BPA: the frame; for frame: the building block number (which consists 1 or more service nodes and compute/storage nodes that are serviced by them - optional).
nodetype
The hardware type of the node. Only can be one of fsp, bpa, cec, frame, ivm, hmc and lpar
supernode
Indicates the connectivity of this CEC in the HFI network. A comma separated list of 2 ids. The first one is the supernode number the CEC is part of. The second one is the logical location number (0-3) of this CEC within the supernode.
sfp
The Service Focal Point of this Frame. This is the name of the HMC that is responsible for collecting hardware service events for this frame and all of the CECs within this frame.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
ppcdirect.5¶
ppcdirect - a table in the xCAT database.
ppcdirect Attributes: hcp, username, password, comments, disable
Info necessary to use FSPs/BPAs to control system p CECs/Frames.
hcp
Hostname of the FSPs/BPAs(for ASMI) and CECs/Frames(for DFM).
username
Userid of the FSP/BPA(for ASMI) and CEC/Frame(for DFM). If not filled in, xCAT will look in the passwd table for key=fsp. If not in the passwd table, the default used is admin.
password
Password of the FSP/BPA(for ASMI) and CEC/Frame(for DFM). If not filled in, xCAT will look in the passwd table for key=fsp. If not in the passwd table, the default used is admin.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
ppchcp.5¶
ppchcp - a table in the xCAT database.
ppchcp Attributes: hcp, username, password, comments, disable
Info necessary to use HMCs and IVMs as hardware control points for LPARs.
hcp
Hostname of the HMC or IVM.
username
Userid of the HMC or IVM. If not filled in, xCAT will look in the passwd table for key=hmc or key=ivm. If not in the passwd table, the default used is hscroot for HMCs and padmin for IVMs.
password
Password of the HMC or IVM. If not filled in, xCAT will look in the passwd table for key=hmc or key=ivm. If not in the passwd table, the default used is abc123 for HMCs and padmin for IVMs.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
prescripts.5¶
prescripts - a table in the xCAT database.
prescripts Attributes: node, begin, end, comments, disable
The scripts that will be run at the beginning and the end of the nodeset(Linux), nimnodeset(AIX) or mkdsklsnode(AIX) command.
node
The node name or group name.
begin
- The scripts to be run at the beginning of the nodeset(Linux),
nimnodeset(AIX) or mkdsklsnode(AIX) command. The format is:
[action1:]s1,s2...[|action2:s3,s4,s5...]
- where:
- action1 and action2 for Linux are the nodeset actions specified in the command. For AIX, action1 and action1 can be ‘diskless’ for mkdsklsnode command’ and ‘standalone for nimnodeset command.
- s1 and s2 are the scripts to run for action1 in order.
- s3, s4, and s5 are the scripts to run for actions2.
If actions are omitted, the scripts apply to all actions. Examples:
myscript1,myscript2 (all actions) diskless:myscript1,myscript2 (AIX) install:myscript1,myscript2|netboot:myscript3 (Linux)All the scripts should be copied to /install/prescripts directory. The following two environment variables will be passed to each script: NODES a coma separated list of node names that need to run the script for ACTION current nodeset action. If '#xCAT setting:MAX_INSTANCE=number' is specified in the script, the script will get invoked for each node in parallel, but no more than number of instances will be invoked at at a time. If it is not specified, the script will be invoked once for all the nodes.
end
- The scripts to be run at the end of the nodeset(Linux),
- nimnodeset(AIX),or mkdsklsnode(AIX) command. The format is the same as the ‘begin’ column.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
prodkey.5¶
prodkey - a table in the xCAT database.
prodkey Attributes: node, product, key, comments, disable
Specify product keys for products that require them
node
The node name or group name.
product
A string to identify the product (for OSes, the osname would be used, i.e. wink28
key
The product key relevant to the aforementioned node/group and product combination
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
rack.5¶
rack - a table in the xCAT database.
rack Attributes: rackname, displayname, num, height, room, comments, disable
Rack information.
rackname
The rack name.
displayname
Alternative name for rack. Only used by PCM.
num
The rack number.
height
Number of units which can be stored in the rack.
room
The room in which the rack is located.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
routes.5¶
routes - a table in the xCAT database.
routes Attributes: routename, net, mask, gateway, ifname, comments, disable
Describes the additional routes needed to be setup in the os routing table. These routes usually are used to connect the management node to the compute node using the servie node as gateway.
routename
Name used to identify this route.
net
The network address.
mask
The network mask.
gateway
The gateway that routes the ip traffic from the mn to the nodes. It is usually a service node.
ifname
The interface name that facing the gateway. It is optional for IPv4 routes, but it is required for IPv6 routes.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
servicenode.5¶
servicenode - a table in the xCAT database.
servicenode Attributes: node, nameserver, dhcpserver, tftpserver, nfsserver, conserver, monserver, ldapserver, ntpserver, ftpserver, nimserver, ipforward, dhcpinterfaces, proxydhcp, comments, disable
List of all Service Nodes and services that will be set up on the Service Node.
node
The hostname of the service node as known by the Management Node.
nameserver
Do we set up DNS on this service node? Valid values: 2, 1, no or 0. If 2, creates named.conf as dns slave, using the management node as dns master, and starts named. If 1, creates named.conf file with forwarding to the management node and starts named. If no or 0, it does not change the current state of the service.
dhcpserver
Do we set up DHCP on this service node? Not supported on AIX. Valid values:yes or 1, no or 0. If yes, runs makedhcp -n. If no or 0, it does not change the current state of the service.
tftpserver
Do we set up TFTP on this service node? Not supported on AIX. Valid values:yes or 1, no or 0. If yes, configures and starts atftp. If no or 0, it does not change the current state of the service.
nfsserver
Do we set up file services (HTTP,FTP,or NFS) on this service node? For AIX will only setup NFS, not HTTP or FTP. Valid values:yes or 1, no or 0.If no or 0, it does not change the current state of the service.
conserver
Do we set up Conserver on this service node? Valid values:yes or 1, no or 0. If yes, configures and starts conserver daemon. If no or 0, it does not change the current state of the service.
monserver
Is this a monitoring event collection point? Valid values:yes or 1, no or 0. If no or 0, it does not change the current state of the service.
ldapserver
Do we set up ldap caching proxy on this service node? Not supported on AIX. Valid values:yes or 1, no or 0. If no or 0, it does not change the current state of the service.
ntpserver
Not used. Use setupntp postscript to setup a ntp server on this service node? Valid values:yes or 1, no or 0. If no or 0, it does not change the current state of the service.
ftpserver
Do we set up a ftp server on this service node? Not supported on AIX Valid values:yes or 1, no or 0. If yes, configure and start vsftpd. (You must manually install vsftpd on the service nodes before this.) If no or 0, it does not change the current state of the service. xCAT is not using ftp for compute nodes provisioning or any other xCAT features, so this attribute can be set to 0 if the ftp service will not be used for other purposes
nimserver
Not used. Do we set up a NIM server on this service node? Valid values:yes or 1, no or 0. If no or 0, it does not change the current state of the service.
ipforward
Do we set up ip forwarding on this service node? Valid values:yes or 1, no or 0. If no or 0, it does not change the current state of the service.
dhcpinterfaces
The network interfaces DHCP server should listen on for the target node. This attribute can be used for management node and service nodes. If defined, it will override the values defined in site.dhcpinterfaces. This is a comma separated list of device names. !remote! indicates a non-local network for relay DHCP. For example: !remote!,eth0,eth1
proxydhcp
Do we set up proxydhcp service on this node? valid values: yes or 1, no or 0. If yes, the proxydhcp daemon will be enabled on this node.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
site.5¶
site - a table in the xCAT database.
site Attributes: key, value, comments, disable
Global settings for the whole cluster. This table is different from the other tables in that each attribute is just named in the key column, rather than having a separate column for each attribute. The following is a list of attributes currently used by xCAT organized into categories.
key
Attribute Name: Description
------------ AIX ATTRIBUTES ------------ nimprime : The name of NIM server, if not set default is the AIX MN. If Linux MN, then must be set for support of mixed cluster (TBD). useSSHonAIX: (yes/1 or no/0). Default is yes. The support for rsh/rcp is deprecated. useNFSv4onAIX: (yes/1 or no/0). If yes, NFSv4 will be used with NIM. If no, NFSv3 will be used with NIM. Default is no. ----------------- DATABASE ATTRIBUTES ----------------- auditnosyslog: If set to 1, then commands will only be written to the auditlog table. This attribute set to 1 and auditskipcmds=ALL means no logging of commands. Default is to write to both the auditlog table and syslog. auditskipcmds: List of commands and/or client types that will not be written to the auditlog table and syslog. See auditnosyslog. 'ALL' means all cmds will be skipped. If attribute is null, all commands will be written. clienttype:web would skip all commands from the web client For example: tabdump,nodels,clienttype:web will not log tabdump,nodels and any web client commands. databaseloc: Directory where we create the db instance directory. Default is /var/lib. Only DB2 is currently supported. Do not use the directory in the site.installloc or installdir attribute. This attribute must not be changed once db2sqlsetup script has been run and DB2 has been setup. excludenodes: A set of comma separated nodes and/or groups that would automatically be subtracted from any noderange, it can be used for excluding some failed nodes for any xCAT commands. See the 'noderange' manpage for details on supported formats. nodestatus: If set to 'n', the nodelist.status column will not be updated during the node deployment, node discovery and power operations. The default is to update. skiptables: Comma separated list of tables to be skipped by dumpxCATdb skipvalidatelog: If set to 1, then getcredentials and getpostscripts calls will not be logged in syslog. ------------- DHCP ATTRIBUTES ------------- dhcpinterfaces: The network interfaces DHCP should listen on. If it is the same for all nodes, use a simple comma-separated list of NICs. To specify different NICs for different nodes: xcatmn|eth1,eth2;service|bond0. In this example xcatmn is the name of the xCAT MN, and DHCP there should listen on eth1 and eth2. On all of the nodes in group 'service' DHCP should listen on the bond0 nic. dhcpsetup: If set to 'n', it will skip the dhcp setup process in the nodeset cmd. dhcplease: The lease time for the dhcp client. The default value is 43200. disjointdhcps: If set to '1', the .leases file on a service node only contains the nodes it manages. The default value is '0'. '0' value means include all the nodes in the subnet. pruneservices: Whether to enable service pruning when noderm is run (i.e. removing DHCP entries when noderm is executed) managedaddressmode: The mode of networking configuration during node provision. If set to 'static', the network configuration will be configured in static mode based on the node and network definition on MN. If set to 'dhcp', the network will be configured with dhcp protocol. The default is 'dhcp'. ------------ DNS ATTRIBUTES ------------ dnshandler: Name of plugin that handles DNS setup for makedns. domain: The DNS domain name used for the cluster. forwarders: The DNS servers at your site that can provide names outside of the cluster. The makedns command will configure the DNS on the management node to forward requests it does not know to these servers. Note that the DNS servers on the service nodes will ignore this value and always be configured to forward requests to the management node. master: The hostname of the xCAT management node, as known by the nodes. nameservers: A comma delimited list of DNS servers that each node in the cluster should use. This value will end up in the nameserver settings of the /etc/resolv.conf on each node. It is common (but not required) to set this attribute value to the IP addr of the xCAT management node, if you have set up the DNS on the management node by running makedns. In a hierarchical cluster, you can also set this attribute to "<xcatmaster>" to mean the DNS server for each node should be the node that is managing it (either its service node or the management node). externaldns: To specify that external dns is used. If externaldns is set to any value then, makedns command will not start the local nameserver on xCAT MN. Default is to start the local nameserver. dnsupdaters: The value are ',' separated string which will be added to the zone config section. This is an interface for user to add configuration entries to the zone sections in named.conf. dnsinterfaces: The network interfaces DNS server should listen on. If it is the same for all nodes, use a simple comma-separated list of NICs. To specify different NICs for different nodes: xcatmn|eth1,eth2;service|bond0. In this example xcatmn is the name of the xCAT MN, and DNS there should listen on eth1 and eth2. On all of the nodes in group 'service' DNS should listen on the bond0 nic. NOTE: if using this attribute to block certain interfaces, make sure the ip maps to your hostname of xCAT MN is not blocked since xCAT needs to use this ip to communicate with the local NDS server on MN. ------------------------- HARDWARE CONTROL ATTRIBUTES ------------------------- blademaxp: The maximum number of concurrent processes for blade hardware control. ea_primary_hmc: The hostname of the HMC that the Integrated Switch Network Management Event Analysis should send hardware serviceable events to for processing and potentially sending to IBM. ea_backup_hmc: The hostname of the HMC that the Integrated Switch Network Management Event Analysis should send hardware serviceable events to if the primary HMC is down. enableASMI: (yes/1 or no/0). If yes, ASMI method will be used after fsp-api. If no, when fsp-api is used, ASMI method will not be used. Default is no. fsptimeout: The timeout, in milliseconds, to use when communicating with FSPs. hwctrldispatch: Whether or not to send hw control operations to the service node of the target nodes. Default is 'y'.(At present, this attribute is only used for IBM Flex System) ipmidispatch: Whether or not to send ipmi hw control operations to the service node of the target compute nodes. Default is 'y'. ipmimaxp: The max # of processes for ipmi hw ctrl. The default is 64. Currently, this is only used for HP hw control. ipmiretries: The # of retries to use when communicating with BMCs. Default is 3. ipmisdrcache: If set to 'no', then the xCAT IPMI support will not cache locally the target node's SDR cache to improve performance. ipmitimeout: The timeout to use when communicating with BMCs. Default is 2. This attribute is currently not used. maxssh: The max # of SSH connections at any one time to the hw ctrl point for PPC This parameter doesn't take effect on the rpower command. It takes effects on other PPC hardware control command getmacs/rnetboot/rbootseq and so on. Default is 8. syspowerinterval: For system p CECs, this is the number of seconds the rpower command will wait between performing the action for each CEC. For system x IPMI servers, this is the number of seconds the rpower command will wait between powering on <syspowermaxnodes> nodes at a time. This value is used to control the power on speed in large clusters. Default is 0. syspowermaxnodes: The number of servers to power on at one time before waiting 'syspowerinterval' seconds to continue on to the next set of nodes. If the noderange given to rpower includes nodes served by different service nodes, it will try to spread each set of nodes across the service nodes evenly. Currently only used for IPMI servers and must be set if 'syspowerinterval' is set. powerinterval: The number of seconds the rpower command to LPARs will wait between performing the action for each LPAR. LPARs of different HCPs (HMCs or FSPs) are done in parallel. This is used to limit the cluster boot up speed in large clusters. Default is 0. This is currently only used for system p hardware. ppcmaxp: The max # of processes for PPC hw ctrl. If there are more than ppcmaxp hcps, this parameter will take effect. It will control the max number of processes for PPC hardware control commands. Default is 64. ppcretry: The max # of PPC hw connection attempts to HMC before failing. It only takes effect on the hardware control commands through HMC. Default is 3. ppctimeout: The timeout, in milliseconds, to use when communicating with PPC hw through HMC. It only takes effect on the hardware control commands through HMC. Default is 0. snmpc: The snmp community string that xcat should use when communicating with the switches. --------------------------- INSTALL/DEPLOYMENT ATTRIBUTES --------------------------- cleanupxcatpost: (yes/1 or no/0). Set to 'yes' or '1' to clean up the /xcatpost directory on the stateless and statelite nodes after the postscripts are run. Default is no. db2installloc: The location which the service nodes should mount for the db2 code to install. Format is hostname:/path. If hostname is omitted, it defaults to the management node. Default is /mntdb2. defserialflow: The default serial flow - currently only used by the mknb command. defserialport: The default serial port - currently only used by mknb. defserialspeed: The default serial speed - currently only used by mknb. genmacprefix: When generating mac addresses automatically, use this manufacturing prefix (e.g. 00:11:aa) genpasswords: Automatically generate random passwords for BMCs when configuring them. installdir: The local directory name used to hold the node deployment packages. installloc: The location from which the service nodes should mount the deployment packages in the format hostname:/path. If hostname is omitted, it defaults to the management node. The path must match the path in the installdir attribute. iscsidir: The path to put the iscsi disks in on the mgmt node. mnroutenames: The name of the routes to be setup on the management node. It is a comma separated list of route names that are defined in the routes table. runbootscripts: If set to 'yes' the scripts listed in the postbootscripts attribute in the osimage and postscripts tables will be run during each reboot of stateful (diskful) nodes. This attribute has no effect on stateless and statelite nodes. Please run the following command after you change the value of this attribute: 'updatenode <nodes> -P setuppostbootscripts' precreatemypostscripts: (yes/1 or no/0). Default is no. If yes, it will instruct xCAT at nodeset and updatenode time to query the db once for all of the nodes passed into the cmd and create the mypostscript file for each node, and put them in a directory of tftpdir(such as: /tftpboot) If no, it will not generate the mypostscript file in the tftpdir. setinstallnic: Set the network configuration for installnic to be static. sharedtftp: Set to 0 or no, xCAT should not assume the directory in tftpdir is mounted on all on Service Nodes. Default is 1/yes. If value is set to a hostname, the directory in tftpdir will be mounted from that hostname on the SN sharedinstall: Indicates if a shared file system will be used for installation resources. Possible values are: 'no', 'sns', or 'all'. 'no' means a shared file system is not being used. 'sns' means a shared filesystem is being used across all service nodes. 'all' means that the management as well as the service nodes are all using a common shared filesystem. The default is 'no'. xcatconfdir: Where xCAT config data is (default /etc/xcat). -------------------- REMOTESHELL ATTRIBUTES -------------------- nodesyncfiledir: The directory on the node, where xdcp will rsync the files SNsyncfiledir: The directory on the Service Node, where xdcp will rsync the files from the MN that will eventually be rsync'd to the compute nodes. sshbetweennodes: Comma separated list of groups of compute nodes to enable passwordless root ssh during install, or xdsh -K. Default is ALLGROUPS. Set to NOGROUPS,if you do not wish to enabled any group of compute nodes. Service Nodes are not affected by this attribute they are always setup with passwordless root access to nodes and other SN. If using the zone table, this attribute in not used. ----------------- SERVICES ATTRIBUTES ----------------- consoleondemand: When set to 'yes', conserver connects and creates the console output only when the user opens the console. Default is no on Linux, yes on AIX. consoleservice: The console service to be used by xCAT. Default is conserver httpport: The port number that the booting/installing nodes should contact the http server on the MN/SN on. It is your responsibility to configure the http server to listen on that port - xCAT will not do that. nmapoptions: Additional options for the nmap command. nmap is used in pping, nodestat, xdsh -v and updatenode commands. Sometimes additional performance tuning may be needed for nmap due to network traffic. For example, if the network response time is too slow, nmap may not give stable output. You can increase the timeout value by specifying '--min-rtt-timeout 1s'. xCAT will append the options defined here to the nmap command. ntpservers: A comma delimited list of NTP servers for the cluster - often the xCAT management node. svloglocal: if set to 1, syslog on the service node will not get forwarded to the mgmt node. timezone: (e.g. America/New_York) tftpdir: tftp directory path. Default is /tftpboot tftpflags: The flags that used to start tftpd. Default is '-v -l -s /tftpboot -m /etc/tftpmapfile4xcat.conf' if tftplfags is not set useNmapfromMN: When set to yes, nodestat command should obtain the node status using nmap (if available) from the management node instead of the service node. This will improve the performance in a flat network. vsftp: Default is 'n'. If set to 'y', the xcatd on the mn will automatically bring up vsftpd. (You must manually install vsftpd before this. This setting does not apply to the service node. For sn you need to set servicenode.ftpserver=1 if you want xcatd to bring up vsftpd. ----------------------- VIRTUALIZATION ATTRIBUTES ----------------------- usexhrm: Have xCAT run its xHRM script when booting up KVM guests to set the virtual network bridge up correctly. See https://sourceforge.net/apps/mediawiki/xcat/index.php?title=XCAT_Virtualization_with_KVM#Setting_up_a_network_bridge vcenterautojoin: When set to no, the VMWare plugin will not attempt to auto remove and add hypervisors while trying to perform operations. If users or tasks outside of xCAT perform the joining this assures xCAT will not interfere. vmwarereconfigonpower: When set to no, the VMWare plugin will make no effort to push vm.cpus/vm.memory updates from xCAT to VMWare. persistkvmguests: Keep the kvm definition on the kvm hypervisor when you power off the kvm guest node. This is useful for you to manually change the kvm xml definition file in virsh for debugging. Set anything means enable. -------------------- XCAT DAEMON ATTRIBUTES -------------------- useflowcontrol: (yes/1 or no/0). If yes, the postscript processing on each node contacts xcatd on the MN/SN using a lightweight UDP packet to wait until xcatd is ready to handle the requests associated with postscripts. This prevents deploying nodes from flooding xcatd and locking out admin interactive use. This value works with the xcatmaxconnections and xcatmaxbatch attributes. Is not supported on AIX. If the value is no, nodes sleep for a random time before contacting xcatd, and retry. The default is no. See the following document for details: Hints_and_Tips_for_Large_Scale_Clusters xcatmaxconnections: Number of concurrent xCAT protocol requests before requests begin queueing. This applies to both client command requests and node requests, e.g. to get postscripts. Default is 64. xcatmaxbatchconnections: Number of concurrent xCAT connections allowed from the nodes. Value must be less than xcatmaxconnections. Default is 50. xcatdport: The port used by the xcatd daemon for client/server communication. xcatiport: The port used by xcatd to receive install status updates from nodes. xcatsslversion: The ssl version by xcatd. Default is SSLv3. xcatsslciphers: The ssl cipher by xcatd. Default is 3DES.
value
The value of the attribute specified in the “key” column.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
statelite.5¶
statelite - a table in the xCAT database.
statelite Attributes: node, image, statemnt, mntopts, comments, disable
The location on an NFS server where a nodes persistent files are stored. Any file marked persistent in the litefile table will be stored in the location specified in this table for that node.
node
The name of the node or group that will use this location.
image
Reserved for future development, not used.
statemnt
The persistant read/write area where a node’s persistent files will be written to, e.g: 10.0.0.1/state/. The node name will be automatically added to the pathname, so 10.0.0.1:/state, will become 10.0.0.1:/state/<nodename>.
mntopts
A comma-separated list of options to use when mounting the persistent directory. (Ex. ‘soft’) The default is to do a ‘hard’ mount.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
storage.5¶
storage - a table in the xCAT database.
storage Attributes: node, osvolume, size, state, storagepool, hypervisor, fcprange, volumetag, type, controller, comments, disable
node
The node name
osvolume
- Specification of what storage to place the node OS image onto. Examples include:
- localdisk (Install to first non-FC attached disk) usbdisk (Install to first USB mass storage device seen) wwn=0x50000393c813840c (Install to storage device with given WWN)
size
Size of the volume. Examples include: 10G, 1024M.
state
State of the volume. The valid values are: free, used, and allocated
storagepool
Name of storage pool where the volume is assigned.
hypervisor
Name of the hypervisor where the volume is configured.
fcprange
A range of acceptable fibre channels that the volume can use. Examples include: 3B00-3C00;4B00-4C00.
volumetag
A specific tag used to identify the volume in the autoyast or kickstart template.
type
The plugin used to drive storage configuration (e.g. svc)
controller
- The management address to attach/detach new volumes.
- In the scenario involving multiple controllers, this data must be passed as argument rather than by table value
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
switch.5¶
switch - a table in the xCAT database.
switch Attributes: node, switch, port, vlan, interface, comments, disable
Contains what switch port numbers each node is connected to.
node
The node name or group name.
switch
The switch hostname.
port
The port number in the switch that this node is connected to. On a simple 1U switch, an administrator can generally enter the number as printed next to the ports, and xCAT will understand switch representation differences. On stacked switches or switches with line cards, administrators should usually use the CLI representation (i.e. 2/0/1 or 5/8). One notable exception is stacked SMC 8848M switches, in which you must add 56 for the proceeding switch, then the port number. For example, port 3 on the second switch in an SMC8848M stack would be 59
vlan
The ID for the tagged vlan that is created on this port using mkvlan and chvlan commands.
interface
The interface name from the node perspective. For example, eth0. For the primary nic, it can be empty, the word “primary” or “primary:ethx” where ethx is the interface name.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
switches.5¶
switches - a table in the xCAT database.
switches Attributes: switch, snmpversion, username, password, privacy, auth, linkports, sshusername, sshpassword, protocol, switchtype, comments, disable
Parameters to use when interrogating switches
switch
The hostname/address of the switch to which the settings apply
snmpversion
The version to use to communicate with switch. SNMPv1 is assumed by default.
username
The username to use for SNMPv3 communication, ignored for SNMPv1
password
The password or community string to use for SNMPv3 or SNMPv1 respectively. Falls back to passwd table, and site snmpc value if using SNMPv1
privacy
The privacy protocol to use for v3. DES is assumed if v3 enabled, as it is the most readily available.
auth
The authentication protocol to use for SNMPv3. SHA is assumed if v3 enabled and this is unspecified
linkports
The ports that connect to other switches. Currently, this column is only used by vlan configuration. The format is: “port_number:switch,port_number:switch...”. Please refer to the switch table for details on how to specify the port numbers.
sshusername
The remote login user name. It can be for ssh or telnet. If it is for telnet, please set protocol to “telnet”.
sshpassword
The remote login password. It can be for ssh or telnet. If it is for telnet, please set protocol to “telnet”.
protocol
Prorocol for running remote commands for the switch. The valid values are: ssh, telnet. ssh is the default. Leave it blank or set to “ssh” for Mellanox IB switch.
switchtype
The type of switch. It is used to identify the file name that implements the functions for this swithc. The valid values are: MellanoxIB etc.
comments
disable
nodels(1), chtab(8), tabdump(8), tabedit(8)
token.5¶
token - a table in the xCAT database.
token Attributes: tokenid, username, expire, comments, disable
The token of users for authentication.
tokenid
It is a UUID as an unified identify for the user.
username
The user name.
expire
The expire time for this token.
comments
Any user-provided notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
virtsd.5¶
virtsd - a table in the xCAT database.
virtsd Attributes: node, sdtype, stype, location, host, cluster, datacenter, comments, disable
The parameters which used to create the Storage Domain
node
The name of the storage domain
sdtype
The type of storage domain. Valid values: data, iso, export
stype
The type of storge. Valid values: nfs, fcp, iscsi, localfs
location
The path of the storage
host
For rhev, a hypervisor host needs to be specified to manage the storage domain as SPM (Storage Pool Manager). But the SPM role will be failed over to another host when this host down.
cluster
A cluster of hosts
datacenter
A collection for all host, vm that will shared the same storages, networks.
comments
disable
nodels(1), chtab(8), tabdump(8), tabedit(8)
vm.5¶
vm - a table in the xCAT database.
vm Attributes: node, mgr, host, migrationdest, storage, storagemodel, storagecache, storageformat, cfgstore, memory, cpus, nics, nicmodel, bootorder, clockoffset, virtflags, master, vncport, textconsole, powerstate, beacon, datacenter, cluster, guestostype, othersettings, physlots, vidmodel, vidproto, vidpassword, comments, disable
Virtualization parameters
node
The node or static group name
mgr
The function manager for the virtual machine
host
The system that currently hosts the VM
migrationdest
A noderange representing candidate destinations for migration (i.e. similar systems, same SAN, or other criteria that xCAT can use
storage
A list of storage files or devices to be used. i.e. dir:///cluster/vm/<nodename> or nfs://<server>/path/to/folder/
storagemodel
Model of storage devices to provide to guest
storagecache
Select caching scheme to employ. E.g. KVM understands ‘none’, ‘writethrough’ and ‘writeback’
storageformat
Select disk format to use by default (e.g. raw versus qcow2)
cfgstore
Optional location for persistant storage separate of emulated hard drives for virtualization solutions that require persistant store to place configuration data
memory
Megabytes of memory the VM currently should be set to.
cpus
Number of CPUs the node should see.
nics
Network configuration parameters. Of the general form [physnet:]interface,.. Generally, interface describes the vlan entity (default for native, tagged for tagged, vl[number] for a specific vlan. physnet is a virtual switch name or port description that is used for some virtualization technologies to construct virtual switches. hypervisor.netmap can map names to hypervisor specific layouts, or the descriptions described there may be used directly here where possible.
nicmodel
Model of NICs that will be provided to VMs (i.e. e1000, rtl8139, virtio, etc)
bootorder
Boot sequence (i.e. net,hd)
clockoffset
Whether to have guest RTC synced to “localtime” or “utc” If not populated, xCAT will guess based on the nodetype.os contents.
virtflags
- General flags used by the virtualization method. For example, in Xen it could, among other things, specify paravirtualized setup, or direct kernel boot. For a hypervisor/dom0 entry, it is the virtualization method (i.e. “xen”). For KVM, the following flag=value pairs are recognized:
- imageformat=[raw|fullraw|qcow2]
- raw is a generic sparse file that allocates storage on demand fullraw is a generic, non-sparse file that preallocates all space qcow2 is a sparse, copy-on-write capable format implemented at the virtualization layer rather than the filesystem level
- clonemethod=[qemu-img|reflink]
- qemu-img allows use of qcow2 to generate virtualization layer copy-on-write reflink uses a generic filesystem facility to clone the files on your behalf, but requires filesystem support such as btrfs
placement_affinity=[migratable|user_migratable|pinned]
master
The name of a master image, if any, this virtual machine is linked to. This is generally set by clonevm and indicates the deletion of a master that would invalidate the storage of this virtual machine
vncport
Tracks the current VNC display port (currently not meant to be set
textconsole
Tracks the Psuedo-TTY that maps to the serial port or console of a VM
powerstate
This flag is used by xCAT to track the last known power state of the VM.
beacon
This flag is used by xCAT to track the state of the identify LED with respect to the VM.
datacenter
Optionally specify a datacenter for the VM to exist in (only applicable to VMWare)
cluster
Specify to the underlying virtualization infrastructure a cluster membership for the hypervisor.
guestostype
This allows administrator to specify an identifier for OS to pass through to virtualization stack. Normally this should be ignored as xCAT will translate from nodetype.os rather than requiring this field be used
othersettings
This allows specifying a semicolon delimited list of key->value pairs to include in a vmx file of VMware. For partitioning on normal power machines, this option is used to specify the hugepage and/or bsr information, the value is like:’hugepage:1,bsr=2’.
physlots
Specify the physical slots drc index that will assigned to the partition, the delimiter is ‘,’, and the drc index must started with ‘0x’. For more details, please reference to manpage of ‘lsvm’.
vidmodel
Model of video adapter to provide to guest. For example, qxl in KVM
vidproto
Request a specific protocol for remote video access be set up. For example, spice in KVM.
vidpassword
Password to use instead of temporary random tokens for VNC and SPICE access
comments
disable
nodels(1), chtab(8), tabdump(8), tabedit(8)
vmmaster.5¶
vmmaster - a table in the xCAT database.
vmmaster Attributes: name, os, arch, profile, storage, storagemodel, nics, vintage, originator, virttype, specializeparameters, comments, disable
Inventory of virtualization images for use with clonevm. Manual intervention in this table is not intended.
name
The name of a master
os
The value of nodetype.os at the time the master was captured
arch
The value of nodetype.arch at the time of capture
profile
The value of nodetype.profile at time of capture
storage
The storage location of bulk master information
storagemodel
The default storage style to use when modifying a vm cloned from this master
nics
The nic configuration and relationship to vlans/bonds/etc
vintage
When this image was created
originator
The user who created the image
virttype
The type of virtualization this image pertains to (e.g. vmware, kvm, etc)
specializeparameters
Implementation specific arguments, currently only “autoLogonCount=<number” for ESXi clonevme
comments
disable
nodels(1), chtab(8), tabdump(8), tabedit(8)
vpd.5¶
vpd - a table in the xCAT database.
vpd Attributes: node, serial, mtm, side, asset, uuid, comments, disable
The Machine type, Model, and Serial numbers of each node.
node
The node name or group name.
serial
The serial number of the node.
mtm
The machine type and model number of the node. E.g. 7984-6BU
side
<BPA>-<port> or <FSP>-<port>. The side information for the BPA/FSP. The side attribute refers to which BPA/FSP, A or B, which is determined by the slot value returned from lsslp command. It also lists the physical port within each BPA/FSP which is determined by the IP address order from the lsslp response. This information is used internally when communicating with the BPAs/FSPs
asset
A field for administators to use to correlate inventory numbers they may have to accomodate
uuid
The UUID applicable to the node
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
websrv.5¶
websrv - a table in the xCAT database.
websrv Attributes: node, port, username, password, comments, disable
Web service parameters
node
The web service hostname.
port
The port of the web service.
username
Userid to use to access the web service.
password
Password to use to access the web service.
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
winimage.5¶
winimage - a table in the xCAT database.
winimage Attributes: imagename, template, installto, partitionfile, winpepath, comments, disable
Information about a Windows operating system image that can be used to deploy cluster nodes.
imagename
The name of this xCAT OS image definition.
template
The fully qualified name of the template file that is used to create the windows unattend.xml file for diskful installation.
installto
The disk and partition that the Windows will be deployed to. The valid format is <disk>:<partition>. If not set, default value is 0:1 for bios boot mode(legacy) and 0:3 for uefi boot mode; If setting to 1, it means 1:1 for bios boot and 1:3 for uefi boot
partitionfile
The path of partition configuration file. Since the partition configuration for bios boot mode and uefi boot mode are different, this configuration file can include both configurations if you need to support both bios and uefi mode. Either way, you must specify the boot mode in the configuration. Example of partition configuration file: [BIOS]xxxxxxx[UEFI]yyyyyyy. To simplify the setting, you also can set installto in partitionfile with section like [INSTALLTO]0:1
winpepath
The path of winpe which will be used to boot this image. If the real path is /tftpboot/winboot/winpe1/, the value for winpepath should be set to winboot/winpe1
comments
Any user-written notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
xcatdb.5¶
An overview of the xCAT database.
The xCAT database contains user settings for the cluster and information gathered from the cluster. It consists of a series of tables, which are described below. To get more information about a particular table, run man for that table name. The tables can be manipulated directly using the tabedit or chtab commands. They can be viewed using nodels or tabdump.
Alternatively, the xCAT database can be viewed and edited as logical objects, instead of flat tables. In this mode, xCAT takes care of which table each attribute should go in. To treat the database as logical object definitions, use the commands: lsdef, mkdef, chdef, rmdef. See Object Definitions below.
xCAT allows the use of different database applications, depending on the needs of your cluster. The default database is SQLite, which is a daemonless, zero-config database. But you could instead choose to use something like postgresql for greater scalability and remote access in the hierarchical/service node case. To use a different database or a different location, create the file /etc/xcat/cfgloc. See the appropriate xCAT docuementation for the format of the file for the database you choose. The following example /etc/xcat/cfgloc file is for PostgreSQL:
Pg:dbname=xcat;host=<mgmtnode>|<pgadminuserid>|<pgadminpasswd>
where mgmtnode is the hostname of the management node adapter on the cluster side, and the pgadminuserid and pgadminpasswd are the database admin and password.
The xCAT database has a number of tables, some with rows that are keyed by node name (such as noderes and nodehm) and others that are not keyed by node name (for example, the policy table). The tables that are keyed by node name have some extra features that enable a more template-based style to be used:
Any group name can be used in lieu of a node name in the node field, and that row will then provide “default” attribute values for any node in that group. A row with a specific node name can then override one or more attribute values for that specific node. For example, if the nodehm table contains:
#node,power,mgt,cons,termserver,termport,conserver,serialport,serialspeed,serialflow,getmac,cmdmapping,comments,disable
"mygroup",,"ipmi",,,,,,"19200",,,,,
"node1",,,,,,,,"115200",,,,,
In the above example, the node group called mygroup sets mgt=ipmi and serialspeed=19200. Any nodes that are in this group will have those attribute values, unless overridden. For example, if node2 is a member of mygroup, it will automatically inherit these attribute values (even though it is not explicitly listed in this table). In the case of node1 above, it inherits mgt=ipmi, but overrides the serialspeed to be 115200, instead of 19200. A useful, typical way to use this capability is to create a node group for your nodes and for all the attribute values that are the same for every node, set them at the group level. Then you only have to set attributes for each node that vary from node to node.
xCAT extends the group capability so that it can also be used for attribute values that vary from node to node in a very regular pattern. For example, if in the ipmi table you want the bmc attribute to be set to whatever the nodename is with “-bmc” appended to the end of it, then use this in the ipmi table:
#node,bmc,bmcport,taggedvlan,bmcid,username,password,comments,disable
"compute","/\z/-bmc/",,,,,,,
In this example, “compute” is a node group that contains all of the compute nodes. The 2nd attribute (bmc) is a regular expression that is similar to a substitution pattern. The 1st part “z” matches the end of the node name and substitutes “-bmc”, effectively appending it to the node name.
Another example is if node1 is to have IP address 10.0.0.1, node2 is to have IP address 10.0.0.2, etc., then this could be represented in the hosts table with the single row:
#node,ip,hostnames,otherinterfaces,comments,disable
"compute","|node(\d+)|10.0.0.($1+0)|",,,,
In this example, the regular expression in the ip attribute uses “|” to separate the 1st and 2nd part. This means that xCAT will allow arithmetic operations in the 2nd part. In the 1st part, “(d+)”, will match the number part of the node name and put that in a variable called $1. The 2nd part is what value to give the ip attribute. In this case it will set it to the string “10.0.0.” and the number that is in $1. (Zero is added to $1 just to remove any leading zeroes.)
A more involved example is with the mp table. If your blades have node names node01, node02, etc., and your chassis node names are cmm01, cmm02, etc., then you might have an mp table like:
#node,mpa,id,nodetype,comments,disable
"blade","|\D+(\d+)|cmm(sprintf('%02d',($1-1)/14+1))|","|\D+(\d+)|(($1-1)%14+1)|",,
Before you panic, let me explain each column:
blade
This is a group name. In this example, we are assuming that all of your blades belong to this group. Each time the xCAT software accesses the mp table to get the management module and slot number of a specific blade (e.g. node20), this row will match (because node20 is in the blade group). Once this row is matched for node20, then the processing described in the following items will take place.
|D+(d+)|cmm(sprintf(‘%02d’,($1-1)/14+1))|
This is a perl substitution pattern that will produce the value for the second column of the table (the management module hostname). The text D+(d+) between the 1st two vertical bars is a regular expression that matches the node name that was searched for in this table (in this example node20). The text that matches within the 1st set of parentheses is set to $1. (If there was a 2nd set of parentheses, it would be set to $2, and so on.) In our case, the D+ matches the non-numeric part of the name (node) and the d+ matches the numeric part (20). So $1 is set to 20. The text cmm(sprintf(‘%02d’,($1-1)/14+1)) between the 2nd and 3rd vertical bars produces the string that should be used as the value for the mpa attribute for node20. Since $1 is set to 20, the expression ($1-1)/14+1 equals 19/14 + 1, which equals 2. (The division is integer division, so 19/14 equals 1. Fourteen is used as the divisor, because there are 14 blades in each chassis.) The value of 2 is then passed into sprintf() with a format string to add a leading zero, if necessary, to always make the number two digits. Lastly the string cmm is added to the beginning, making the resulting string cmm02, which will be used as the hostname of the management module.
|D+(d+)|(($1-1)%14+1)|
This item is similar to the one above. This substituion pattern will produce the value for the 3rd column (the chassis slot number for this blade). Because this row was the match for node20, the parentheses within the 1st set of vertical bars will set $1 to 20. Since % means modulo division, the expression ($1-1)%14+1 will evaluate to 6.
See http://www.perl.com/doc/manual/html/pod/perlre.html for information on perl regular expressions.
As of xCAT 2.8.1, you can use a modified version of the regular expression support described in the previous section. You do not need to enter the node information (1st part of the expression), it will be derived from the input nodename. You only need to supply the 2nd part of the expression to determine the value to give the attribute. For examples, see
https://sourceforge.net/p/xcat/wiki/Listing_and_Modifying_the_Database/#easy-regular-expressions
Because it can get confusing what attributes need to go in what tables, the xCAT database can also be viewed and edited as logical objects, instead of flat tables. Use mkdef, chdef, lsdef, and rmdef to create, change, list, and delete objects. When using these commands, the object attributes will be stored in the same tables, as if you edited the tables by hand. The only difference is that the object commands take care of knowing which tables all of the information should go in.
To run man for any of the object definitions below, use section 7. For example: man 7 node
The object types are:
auditlog(7)|auditlog.7
boottarget(7)|boottarget.7
eventlog(7)|eventlog.7
firmware(7)|firmware.7
group(7)|group.7
kit(7)|kit.7
kitcomponent(7)|kitcomponent.7
kitrepo(7)|kitrepo.7
monitoring(7)|monitoring.7
network(7)|network.7
node(7)|node.7
notification(7)|notification.7
osdistro(7)|osdistro.7
osdistroupdate(7)|osdistroupdate.7
osimage(7)|osimage.7
policy(7)|policy.7
rack(7)|rack.7
route(7)|route.7
site(7)|site.7
zone(7)|zone.7
To manipulate the tables directly, use nodels(1), chtab(8), tabdump(8), tabedit(8), nodeadd(8), nodech(1).
To run man for any of the table descriptions below, use section 5. For example: man 5 nodehm
The tables are:
auditlog(5)|auditlog.5
Audit Data log.
bootparams(5)|bootparams.5
Current boot settings to be sent to systems attempting network boot for deployment, stateless, or other reasons. Mostly automatically manipulated by xCAT.
boottarget(5)|boottarget.5
Specify non-standard initrd, kernel, and parameters that should be used for a given profile.
cfgmgt(5)|cfgmgt.5
Configuration management data for nodes used by non-xCAT osimage management services to install and configure software on a node.
chain(5)|chain.5
Controls what operations are done (and it what order) when a node is discovered and deployed.
deps(5)|deps.5
Describes dependencies some nodes have on others. This can be used, e.g., by rpower -d to power nodes on or off in the correct order.
discoverydata(5)|discoverydata.5
Discovery data which sent from genesis.
domain(5)|domain.5
Mapping of nodes to domain attributes
eventlog(5)|eventlog.5
Stores the events occurred.
firmware(5)|firmware.5
Maps node to firmware values to be used for setup at node discovery or later
hosts(5)|hosts.5
IP addresses and hostnames of nodes. This info is optional and is only used to populate /etc/hosts and DNS via makehosts and makedns. Using regular expressions in this table can be a quick way to populate /etc/hosts.
hwinv(5)|hwinv.5
The hareware inventory for the node.
hypervisor(5)|hypervisor.5
Hypervisor parameters
ipmi(5)|ipmi.5
Settings for nodes that are controlled by an on-board BMC via IPMI.
iscsi(5)|iscsi.5
Contains settings that control how to boot a node from an iSCSI target
kit(5)|kit.5
This table stores all kits added to the xCAT cluster.
kitcomponent(5)|kitcomponent.5
This table stores all kit components added to the xCAT cluster.
kitrepo(5)|kitrepo.5
This table stores all kits added to the xCAT cluster.
kvm_masterdata(5)|kvm_masterdata.5
Persistant store for KVM plugin for masters
kvm_nodedata(5)|kvm_nodedata.5
Persistant store for KVM plugin, not intended for manual modification.
linuximage(5)|linuximage.5
Information about a Linux operating system image that can be used to deploy cluster nodes.
litefile(5)|litefile.5
The litefile table specifies the directories and files on the statelite nodes that should be readwrite, persistent, or readonly overlay. All other files in the statelite nodes come from the readonly statelite image.
litetree(5)|litetree.5
Directory hierarchy to traverse to get the initial contents of node files. The files that are specified in the litefile table are searched for in the directories specified in this table.
mac(5)|mac.5
The MAC address of the node’s install adapter. Normally this table is populated by getmacs or node discovery, but you can also add entries to it manually.
mic(5)|mic.5
The host, slot id and configuraton of the mic (Many Integrated Core).
monitoring(5)|monitoring.5
Controls what external monitoring tools xCAT sets up and uses. Entries should be added and removed from this table using the provided xCAT commands monstart and monstop.
monsetting(5)|monsetting.5
Specifies the monitoring plug-in specific settings. These settings will be used by the monitoring plug-in to customize the behavior such as event filter, sample interval, responses etc. Entries should be added, removed or modified by chtab command. Entries can also be added or modified by the monstart command when a monitoring plug-in is brought up.
mp(5)|mp.5
Contains the hardware control info specific to blades. This table also refers to the mpa table, which contains info about each Management Module.
mpa(5)|mpa.5
Contains info about each Management Module and how to access it.
networks(5)|networks.5
Describes the networks in the cluster and info necessary to set up nodes on that network.
nics(5)|nics.5
Stores NIC details.
nimimage(5)|nimimage.5
All the info that specifies a particular AIX operating system image that can be used to deploy AIX nodes.
nodegroup(5)|nodegroup.5
Contains group definitions, whose membership is dynamic depending on characteristics of the node.
nodehm(5)|nodehm.5
Settings that control how each node’s hardware is managed. Typically, an additional table that is specific to the hardware type of the node contains additional info. E.g. the ipmi, mp, and ppc tables.
nodelist(5)|nodelist.5
The list of all the nodes in the cluster, including each node’s current status and what groups it is in.
nodepos(5)|nodepos.5
Contains info about the physical location of each node. Currently, this info is not used by xCAT, and therefore can be in whatevery format you want. It will likely be used in xCAT in the future.
noderes(5)|noderes.5
Resources and settings to use when installing nodes.
nodetype(5)|nodetype.5
A few hardware and software characteristics of the nodes.
notification(5)|notification.5
Contains registrations to be notified when a table in the xCAT database changes. Users can add entries to have additional software notified of changes. Add and remove entries using the provided xCAT commands regnotif and unregnotif.
osdistro(5)|osdistro.5
Information about all the OS distros in the xCAT cluster
osdistroupdate(5)|osdistroupdate.5
Information about the OS distro updates in the xCAT cluster
osimage(5)|osimage.5
Basic information about an operating system image that can be used to deploy cluster nodes.
passwd(5)|passwd.5
Contains default userids and passwords for xCAT to access cluster components. In most cases, xCAT will also actually set the userid/password in the relevant component when it is being configured or installed. Userids/passwords for specific cluster components can be overidden in other tables, e.g. mpa, ipmi, ppchcp, etc.
performance(5)|performance.5
Describes the system performance every interval unit of time.
policy(5)|policy.5
The policy table in the xCAT database controls who has authority to run specific xCAT operations. It is basically the Access Control List (ACL) for xCAT. It is sorted on the priority field before evaluating.
postscripts(5)|postscripts.5
The scripts that should be run on each node after installation or diskless boot.
ppc(5)|ppc.5
List of system p hardware: HMCs, IVMs, FSPs, BPCs, CECs, Frames.
ppcdirect(5)|ppcdirect.5
Info necessary to use FSPs/BPAs to control system p CECs/Frames.
ppchcp(5)|ppchcp.5
Info necessary to use HMCs and IVMs as hardware control points for LPARs.
prescripts(5)|prescripts.5
The scripts that will be run at the beginning and the end of the nodeset(Linux), nimnodeset(AIX) or mkdsklsnode(AIX) command.
prodkey(5)|prodkey.5
Specify product keys for products that require them
rack(5)|rack.5
Rack information.
routes(5)|routes.5
Describes the additional routes needed to be setup in the os routing table. These routes usually are used to connect the management node to the compute node using the servie node as gateway.
servicenode(5)|servicenode.5
List of all Service Nodes and services that will be set up on the Service Node.
site(5)|site.5
Global settings for the whole cluster. This table is different from the other tables in that each attribute is just named in the key column, rather than having a separate column for each attribute. The following is a list of attributes currently used by xCAT organized into categories.
statelite(5)|statelite.5
The location on an NFS server where a nodes persistent files are stored. Any file marked persistent in the litefile table will be stored in the location specified in this table for that node.
storage(5)|storage.5
switch(5)|switch.5
Contains what switch port numbers each node is connected to.
switches(5)|switches.5
Parameters to use when interrogating switches
token(5)|token.5
The token of users for authentication.
virtsd(5)|virtsd.5
The parameters which used to create the Storage Domain
vm(5)|vm.5
Virtualization parameters
vmmaster(5)|vmmaster.5
Inventory of virtualization images for use with clonevm. Manual intervention in this table is not intended.
vpd(5)|vpd.5
The Machine type, Model, and Serial numbers of each node.
websrv(5)|websrv.5
Web service parameters
winimage(5)|winimage.5
Information about a Windows operating system image that can be used to deploy cluster nodes.
zone(5)|zone.5
Defines a cluster zone for nodes that share root ssh key access to each other.
zvm(5)|zvm.5
List of z/VM virtual servers.
nodels(1), chtab(8), tabdump(8), tabedit(8), lsdef(1), mkdef(1), chdef(1), rmdef(1)
xcatstanzafile.5¶
xcatstanzafile - Format of a stanza file that can be used with xCAT data object definition commands.
A stanza file contains information that can be used to create xCAT data object definitions. A stanza file can be used as input to several xCAT commands. The stanza file contains one or more individual stanzas that provide information for individual object definitions. The following rules must be followed when creating a stanza file:
*
An object stanza header consists of the object name followed by a colon, (”:”).
*
Attribute lines must take the form of Attribute=Value.
*
Attribute name might include the character dot (”.”), like passwd.HMC and nicips.eth0.
*
Only one stanza can exist for each object name.
*
All stanzas except for default stanzas must have a value set for “objtype”.
*
Comments beginning with the “#” pound sign may be added to the file. A comment must be on a separate line.
*
When parsing the file, tab characters and spaces are ignored.
*
Each line of the file can have no more than one header or attribute definition.
*
If the header name is “default-<object type>:” the attribute values in the stanza are considered default values for subsequent definitions in the file that are the same object type.
*
Default stanzas can be specified multiple times and at any point in a stanza file. The values apply to all definitions following the default stanzas in a file. The default values are cumulative; a default attribute value will remain set until it is explicitly unset or changed.
*
To turn off a default value, use another default stanza to set the attribute to have no value using a blank space.
*
When a specific value for an attribute is provided in the stanza, it takes priority over any default value that had been set.
The format of a stanza file should look similar to the following.
default-<object type>:
attr=val
attr=val
. . .
<object name>:
objtype=<object type>
attr=val
attr=val
. . .
<object name>:
objtype=<object type>
attr=val
attr=val
. . .
Sample stanza file:
mysite: objtype=site rsh=/bin/rsh rcp=/bin/rcp installdir=/xcatinstall domain=ppd.pok.ibm.com MSnet01: objtype=network gateway=1.2.3.4 netmask=255.255.255.0 nameserver=5.6.7.8 default-node: next_osimage=aix61 network=MSnet01 groups=all,compute node01: objtype=node MAC=A2E26002C003 xcatmaster=MS02.ppd.pok.com nfsserver=IS227.ppd.pok.com node02: objtype=node MAC=A2E26002B004 xcatmaster=MS01.ppd.pok.com nfsserver=IS127.ppd.pok.com grp01: objtype=group members=node1,node2,node3
This file is part of xCAT software product.
mkdef(1)|mkdef.1, lsdef(1)|lsdef.1, rmdef(1)|rmdef.1, chdef(1)|chdef.1
zone.5¶
zone - a table in the xCAT database.
zone Attributes: zonename, sshkeydir, sshbetweennodes, defaultzone, comments, disable
Defines a cluster zone for nodes that share root ssh key access to each other.
zonename
The name of the zone.
sshkeydir
Directory containing the shared root ssh RSA keys.
sshbetweennodes
Indicates whether passwordless ssh will be setup between the nodes of this zone. Values are yes/1 or no/0. Default is yes.
defaultzone
If nodes are not assigned to any other zone, they will default to this zone. If value is set to yes or 1.
comments
Any user-provided notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
zvm.5¶
zvm - a table in the xCAT database.
zvm Attributes: node, hcp, userid, nodetype, parent, comments, disable
List of z/VM virtual servers.
node
The node name.
hcp
The hardware control point for this node.
userid
The z/VM userID of this node.
nodetype
The node type. Valid values: cec (Central Electronic Complex), lpar (logical partition), zvm (z/VM host operating system), and vm (virtual machine).
parent
The parent node. For LPAR, this specifies the CEC. For z/VM, this specifies the LPAR. For VM, this specifies the z/VM host operating system.
comments
Any user provided notes.
disable
Set to ‘yes’ or ‘1’ to comment out this row.
nodels(1), chtab(8), tabdump(8), tabedit(8)
man7¶
auditlog.7¶
auditlog - a logical object definition in the xCAT database.
auditlog Attributes: args, audittime, clientname, clienttype, command, comments, disable, noderange, recid, status, userid
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
args (auditlog.args)
The command argument list.
audittime (auditlog.audittime)
The timestamp for the audit entry.
clientname (auditlog.clientname)
The client machine, where the command originated.
clienttype (auditlog.clienttype)
Type of command: cli,java,webui,other.
command (auditlog.command)
Command executed.
comments (auditlog.comments)
Any user-provided notes.
disable (auditlog.disable)
Do not use. tabprune will not work if set to yes or 1
noderange (auditlog.noderange)
The noderange on which the command was run.
recid (auditlog.recid)
The record id.
status (auditlog.status)
Allowed or Denied.
userid (auditlog.userid)
The user running the command.
mkdef(1), chdef(1), lsdef(1), rmdef(1)
boottarget.7¶
boottarget - a logical object definition in the xCAT database.
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
mkdef(1), chdef(1), lsdef(1), rmdef(1)
eventlog.7¶
eventlog - a logical object definition in the xCAT database.
eventlog Attributes: application, comments, component, disable, eventtime, eventtype, id, message, monitor, monnode, node, rawdata, recid, severity
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
application (eventlog.application)
The application that reports the event.
comments (eventlog.comments)
Any user-provided notes.
component (eventlog.component)
The component where the event occurred.
disable (eventlog.disable)
Do not use. tabprune will not work if set to yes or 1
eventtime (eventlog.eventtime)
The timestamp for the event.
eventtype (eventlog.eventtype)
The type of the event.
id (eventlog.id)
The location or the resource name where the event occurred.
message (eventlog.message)
The full description of the event.
monitor (eventlog.monitor)
The name of the monitor that monitors this event.
monnode (eventlog.monnode)
The node that monitors this event.
node (eventlog.node)
The node where the event occurred.
rawdata (eventlog.rawdata)
The data that associated with the event.
recid (eventlog.recid)
The record id.
severity (eventlog.severity)
The severity of the event. Valid values are: informational, warning, critical.
mkdef(1), chdef(1), lsdef(1), rmdef(1)
firmware.7¶
firmware - a logical object definition in the xCAT database.
firmware Attributes: cfgfile, comments, disable
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
cfgfile (firmware.cfgfile)
The file to use.
comments (firmware.comments)
Any user-written notes.
disable (firmware.disable)
Set to ‘yes’ or ‘1’ to comment out this row.
mkdef(1), chdef(1), lsdef(1), rmdef(1)
group.7¶
group - a logical object definition in the xCAT database.
group Attributes: addkcmdline, arch, authdomain, bmc, bmcpassword, bmcport, bmcusername, cfgmgr, cfgmgtroles, cfgserver, chain, chassis, cmdmapping, cons, conserver, consoleondemand, cpucount, cputype, currchain, currstate, dhcpinterfaces, disksize, displayname, domainadminpassword, domainadminuser, domaintype, getmac, groupname, grouptype, hcp, height, hostcluster, hostinterface, hostmanager, hostnames, hosttype, hwtype, id, initrd, installnic, interface, ip, iscsipassword, iscsiserver, iscsitarget, iscsiuserid, kcmdline, kernel, mac, membergroups, members, memory, mgt, micbridge, michost, micid, miconboot, micpowermgt, micvlog, migrationdest, monserver, mpa, mtm, nameservers, netboot, nfsdir, nfsserver, nicaliases, niccustomscripts, nicextraparams, nichostnameprefixes, nichostnamesuffixes, nicips, nicnetworks, nictypes, nimserver, nodetype, ondiscover, os, osvolume, otherinterfaces, ou, parent, passwd.HMC, passwd.admin, passwd.celogin, passwd.general, passwd.hscroot, password, postbootscripts, postscripts, power, pprofile, prescripts-begin, prescripts-end, primarynic, productkey, profile, provmethod, rack, room, routenames, serial, serialflow, serialport, serialspeed, servicenode, setupconserver, setupdhcp, setupftp, setupipforward, setupldap, setupnameserver, setupnfs, setupnim, setupntp, setupproxydhcp, setuptftp, sfp, side, slot, slotid, slots, storagcontroller, storagetype, supernode, supportedarchs, supportproxydhcp, switch, switchinterface, switchport, switchvlan, termport, termserver, tftpdir, tftpserver, unit, urlpath, usercomment, userid, username, vmbeacon, vmbootorder, vmcfgstore, vmcluster, vmcpus, vmhost, vmmanager, vmmaster, vmmemory, vmnicnicmodel, vmnics, vmothersetting, vmphyslots, vmstorage, vmstoragecache, vmstorageformat, vmstoragemodel, vmtextconsole, vmvirtflags, vmvncport, webport, wherevals, xcatmaster
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
addkcmdline (bootparams.addkcmdline)
User specified one or more parameters to be passed to the kernel. For the kernel options need to be persistent after installation, specify them with prefix “R::”
arch (nodetype.arch)
The hardware architecture of this node. Valid values: x86_64, ppc64, x86, ia64.
authdomain (domain.authdomain)
If a node should participate in an AD domain or Kerberos realm distinct from domain indicated in site, this field can be used to specify that
bmc (ipmi.bmc)
The hostname of the BMC adapater.
bmcpassword (ipmi.password)
The BMC password. If not specified, the key=ipmi row in the passwd table is used as the default.
bmcport (ipmi.bmcport)
In systems with selectable shared/dedicated ethernet ports, this parameter can be used to specify the preferred port. 0 means use the shared port, 1 means dedicated, blank is to not assign. The following special cases exist for IBM System x servers: For x3755 M3 systems, 0 means use the dedicated port, 1 means shared, blank is to not assign. For certain systems which have a mezzaine or ML2 adapter, there is a second value to include: For x3750 M4 (Model 8722): 0 2 1st 1Gbps interface for LOM 0 0 1st 10Gbps interface for LOM 0 3 2nd 1Gbps interface for LOM 0 1 2nd 10Gbps interface for LOM For x3750 M4 (Model 8752), x3850/3950 X6, dx360 M4, x3550 M4, and x3650 M4: 0 Shared (1st onboard interface) 1 Dedicated 2 0 First interface on ML2 or mezzanine adapter 2 1 Second interface on ML2 or mezzanine adapter 2 2 Third interface on ML2 or mezzanine adapter 2 3 Fourth interface on ML2 or mezzanine adapter
bmcusername (ipmi.username)
The BMC userid. If not specified, the key=ipmi row in the passwd table is used as the default.
cfgmgr (cfgmgt.cfgmgr)
The name of the configuration manager service. Currently ‘chef’ and ‘puppet’ are supported services.
cfgmgtroles (cfgmgt.roles)
The roles associated with this node as recognized by the cfgmgr for the software that is to be installed and configured. These role names map to chef recipes or puppet manifest classes that should be used for this node. For example, chef OpenStack cookbooks have roles such as mysql-master,keystone, glance, nova-controller, nova-conductor, cinder-all.
cfgserver (cfgmgt.cfgserver)
The xCAT node name of the chef server or puppet master
chain (chain.chain)
A comma-delimited chain of actions to be performed automatically when this node is discovered. (“Discovered” means a node booted, but xCAT and DHCP did not recognize the MAC of this node. In this situation, xCAT initiates the discovery process, the last step of which is to run the operations listed in this chain attribute, one by one.) Valid values: boot or reboot, install or netboot, runcmd=<cmd>, runimage=<URL>, shell, standby. (Default - same as no chain - it will do only the discovery.). Example, for BMC machines use: runcmd=bmcsetup,shell.
chassis (nodepos.chassis)
The BladeCenter chassis the blade is in.
cmdmapping (nodehm.cmdmapping)
The fully qualified name of the file that stores the mapping between PCM hardware management commands and xCAT/third-party hardware management commands for a particular type of hardware device. Only used by PCM.
cons (nodehm.cons)
The console method. If nodehm.serialport is set, this will default to the nodehm.mgt setting, otherwise it defaults to unused. Valid values: cyclades, mrv, or the values valid for the mgt attribute.
conserver (nodehm.conserver)
The hostname of the machine where the conserver daemon is running. If not set, the default is the xCAT management node.
consoleondemand (nodehm.consoleondemand)
This overrides the value from site.consoleondemand; (0=no, 1=yes). Default is the result from site.consoleondemand.
cpucount (hwinv.cpucount)
The number of cpus for the node.
cputype (hwinv.cputype)
The cpu model name for the node.
currchain (chain.currchain)
The chain steps still left to do for this node. This attribute will be automatically adjusted by xCAT while xCAT-genesis is running on the node (either during node discovery or a special operation like firmware update). During node discovery, this attribute is initialized from the chain attribute and updated as the chain steps are executed.
currstate (chain.currstate)
The current or next chain step to be executed on this node by xCAT-genesis. Set by xCAT during node discovery or as a result of nodeset.
dhcpinterfaces (servicenode.dhcpinterfaces)
The network interfaces DHCP server should listen on for the target node. This attribute can be used for management node and service nodes. If defined, it will override the values defined in site.dhcpinterfaces. This is a comma separated list of device names. !remote! indicates a non-local network for relay DHCP. For example: !remote!,eth0,eth1
disksize (hwinv.disksize)
The size of the disks for the node in GB.
displayname (mpa.displayname)
Alternative name for BladeCenter chassis. Only used by PCM.
domainadminpassword (domain.adminpassword)
Allow a node specific indication of Administrative user password for the domain. Most will want to ignore this in favor of passwd table.
domainadminuser (domain.adminuser)
Allow a node specific indication of Administrative user. Most will want to just use passwd table to indicate this once rather than by node.
domaintype (domain.type)
Type, if any, of authentication domain to manipulate. The only recognized value at the moment is activedirectory.
getmac (nodehm.getmac)
The method to use to get MAC address of the node with the getmac command. If not set, the mgt attribute will be used. Valid values: same as values for mgmt attribute.
groupname (nodegroup.groupname)
Name of the group.
grouptype (nodegroup.grouptype)
The only current valid value is dynamic. We will be looking at having the object def commands working with static group definitions in the nodelist table.
hcp (ppc.hcp, zvm.hcp)
The hardware control point for this node (HMC, IVM, Frame or CEC). Do not need to set for BPAs and FSPs.
or
The hardware control point for this node.
height (nodepos.height)
The server height in U(s).
hostcluster (hypervisor.cluster)
Specify to the underlying virtualization infrastructure a cluster membership for the hypervisor.
hostinterface (hypervisor.interface)
The definition of interfaces for the hypervisor. The format is [networkname:interfacename:bootprotocol:IP:netmask:gateway] that split with | for each interface
hostmanager (hypervisor.mgr)
The virtualization specific manager of this hypervisor when applicable
hostnames (hosts.hostnames)
Hostname aliases added to /etc/hosts for this node. Comma or blank separated list.
hosttype (hypervisor.type)
The plugin associated with hypervisor specific commands such as revacuate
hwtype (ppc.nodetype, zvm.nodetype, mp.nodetype, mic.nodetype)
The hardware type of the node. Only can be one of fsp, bpa, cec, frame, ivm, hmc and lpar
or
The node type. Valid values: cec (Central Electronic Complex), lpar (logical partition), zvm (z/VM host operating system), and vm (virtual machine).
or
The hardware type for mp node. Valid values: mm,cmm, blade.
or
The hardware type of the mic node. Generally, it is mic.
id (ppc.id, mp.id)
For LPARs: the LPAR numeric id; for CECs: the cage number; for Frames: the frame number.
or
The slot number of this blade in the BladeCenter chassis.
initrd (bootparams.initrd)
The initial ramdisk image that network boot actions should use (could be a DOS floppy or hard drive image if using memdisk as kernel)
installnic (noderes.installnic)
The network adapter on the node that will be used for OS deployment, the installnic can be set to the network adapter name or the mac address or the keyword “mac” which means that the network interface specified by the mac address in the mac table will be used. If not set, primarynic will be used. If primarynic is not set too, the keyword “mac” will be used as default.
interface (mac.interface)
The adapter interface name that will be used to install and manage the node. E.g. eth0 (for linux) or en0 (for AIX).)
ip (hosts.ip)
The IP address of the node. This is only used in makehosts. The rest of xCAT uses system name resolution to resolve node names to IP addresses.
iscsipassword (iscsi.passwd)
The password for the iscsi server containing the boot device for this node.
iscsiserver (iscsi.server)
The server containing the iscsi boot device for this node.
iscsitarget (iscsi.target)
The iscsi disk used for the boot device for this node. Filled in by xCAT.
iscsiuserid (iscsi.userid)
The userid of the iscsi server containing the boot device for this node.
kcmdline (bootparams.kcmdline)
Arguments to be passed to the kernel
kernel (bootparams.kernel)
The kernel that network boot actions should currently acquire and use. Note this could be a chained boot loader such as memdisk or a non-linux boot loader
mac (mac.mac)
The mac address or addresses for which xCAT will manage static bindings for this node. This may be simply a mac address, which would be bound to the node name (such as “01:02:03:04:05:0E”). This may also be a “|” delimited string of “mac address!hostname” format (such as “01:02:03:04:05:0E!node5|01:02:03:05:0F!node6-eth1”).
membergroups (nodegroup.membergroups)
This attribute stores a comma-separated list of nodegroups that this nodegroup refers to. This attribute is only used by PCM.
members (nodegroup.members)
The value of the attribute is not used, but the attribute is necessary as a place holder for the object def commands. (The membership for static groups is stored in the nodelist table.)
memory (hwinv.memory)
The size of the memory for the node in MB.
mgt (nodehm.mgt)
The method to use to do general hardware management of the node. This attribute is used as the default if power or getmac is not set. Valid values: ipmi, blade, hmc, ivm, fsp, bpa, kvm, esx, rhevm. See the power attribute for more details.
micbridge (mic.bridge)
The virtual bridge on the host node which the mic connected to.
michost (mic.host)
The host node which the mic card installed on.
micid (mic.id)
The device id of the mic node.
miconboot (mic.onboot)
Set mic to autoboot when mpss start. Valid values: yes|no. Default is yes.
micpowermgt (mic.powermgt)
Set the Power Management for mic node. This attribute is used to set the power management state that mic may get into when it is idle. Four states can be set: cpufreq, corec6, pc3 and pc6. The valid value for powermgt attribute should be [cpufreq=<on|off>]![corec6=<on|off>]![pc3=<on|off>]![pc6=<on|off>]. e.g. cpufreq=on!corec6=off!pc3=on!pc6=off. Refer to the doc of mic to get more information for power management.
micvlog (mic.vlog)
Set the Verbose Log to console. Valid values: yes|no. Default is no.
migrationdest (vm.migrationdest)
A noderange representing candidate destinations for migration (i.e. similar systems, same SAN, or other criteria that xCAT can use
monserver (noderes.monserver)
The monitoring aggregation point for this node. The format is “x,y” where x is the ip address as known by the management node and y is the ip address as known by the node.
mpa (mp.mpa)
The managment module used to control this blade.
mtm (vpd.mtm)
The machine type and model number of the node. E.g. 7984-6BU
nameservers (noderes.nameservers)
An optional node/group specific override for name server list. Most people want to stick to site or network defined nameserver configuration.
netboot (noderes.netboot)
The type of network booting to use for this node. Valid values: pxe or xnba for x86* architecture, yaboot for POWER architecture, grub2-tftp and grub2-http for RHEL7 on Power and all the os deployment on Power LE. Notice: yaboot is not supported from rhels7 on Power,use grub2-tftp or grub2-http instead, the difference between the 2 is the file transfer protocol(i.e, http or tftp)
nfsdir (noderes.nfsdir)
The path that should be mounted from the NFS server.
nfsserver (noderes.nfsserver)
The NFS or HTTP server for this node (as known by this node).
nicaliases (nics.nicaliases)
Comma-separated list of hostname aliases for each NIC.
Format: eth0!<alias list>,eth1!<alias1 list>|<alias2 list> For multiple aliases per nic use a space-separated list. For example: eth0!moe larry curly,eth1!tom|jerry
niccustomscripts (nics.niccustomscripts)
Comma-separated list of custom scripts per NIC. <nic1>!<script1>,<nic2>!<script2>, e.g. eth0!configeth eth0, ib0!configib ib0. The xCAT object definition commands support to use niccustomscripts.<nicname> as the sub attribute
.
nicextraparams (nics.nicextraparams)
Comma-separated list of extra parameters that will be used for each NIC configuration.
If only one ip address is associated with each NIC: <nic1>!<param1=value1 param2=value2>,<nic2>!<param3=value3>, for example, eth0!MTU=1500,ib0!MTU=65520 CONNECTED_MODE=yes. If multiple ip addresses are associated with each NIC: <nic1>!<param1=value1 param2=value2>|<param3=value3>,<nic2>!<param4=value4 param5=value5>|<param6=value6>, for example, eth0!MTU=1500|MTU=1460,ib0!MTU=65520 CONNECTED_MODE=yes. The xCAT object definition commands support to use nicextraparams.<nicname> as the sub attributes.
nichostnameprefixes (nics.nichostnameprefixes)
Comma-separated list of hostname prefixes per NIC.
If only one ip address is associated with each NIC: <nic1>!<ext1>,<nic2>!<ext2>,..., for example, eth0!eth0-,ib0!ib- If multiple ip addresses are associcated with each NIC: <nic1>!<ext1>|<ext2>,<nic2>!<ext1>|<ext2>,..., for example, eth0!eth0-|eth0-ipv6i-,ib0!ib-|ib-ipv6-. The xCAT object definition commands support to use nichostnameprefixes.<nicname> as the sub attributes. Note: According to DNS rules a hostname must be a text string up to 24 characters drawn from the alphabet (A-Z), digits (0-9), minus sign (-),and period (.). When you are specifying "nichostnameprefixes" or "nicaliases" make sure the resulting hostnames will conform to this naming convention
nichostnamesuffixes (nics.nichostnamesuffixes)
Comma-separated list of hostname suffixes per NIC.
If only one ip address is associated with each NIC: <nic1>!<ext1>,<nic2>!<ext2>,..., for example, eth0!-eth0,ib0!-ib0 If multiple ip addresses are associcated with each NIC: <nic1>!<ext1>|<ext2>,<nic2>!<ext1>|<ext2>,..., for example, eth0!-eth0|-eth0-ipv6,ib0!-ib0|-ib0-ipv6. The xCAT object definition commands support to use nichostnamesuffixes.<nicname> as the sub attributes. Note: According to DNS rules a hostname must be a text string up to 24 characters drawn from the alphabet (A-Z), digits (0-9), minus sign (-),and period (.). When you are specifying "nichostnamesuffixes" or "nicaliases" make sure the resulting hostnames will conform to this naming convention
nicips (nics.nicips)
Comma-separated list of IP addresses per NIC. To specify one ip address per NIC:
<nic1>!<ip1>,<nic2>!<ip2>,..., for example, eth0!10.0.0.100,ib0!11.0.0.100 To specify multiple ip addresses per NIC: <nic1>!<ip1>|<ip2>,<nic2>!<ip1>|<ip2>,..., for example, eth0!10.0.0.100|fd55::214:5eff:fe15:849b,ib0!11.0.0.100|2001::214:5eff:fe15:849a. The xCAT object definition commands support to use nicips.<nicname> as the sub attributes. Note: The primary IP address must also be stored in the hosts.ip attribute. The nichostnamesuffixes should specify one hostname suffix for each ip address.
nicnetworks (nics.nicnetworks)
Comma-separated list of networks connected to each NIC.
If only one ip address is associated with each NIC: <nic1>!<network1>,<nic2>!<network2>, for example, eth0!10_0_0_0-255_255_0_0, ib0!11_0_0_0-255_255_0_0 If multiple ip addresses are associated with each NIC: <nic1>!<network1>|<network2>,<nic2>!<network1>|<network2>, for example, eth0!10_0_0_0-255_255_0_0|fd55:faaf:e1ab:336::/64,ib0!11_0_0_0-255_255_0_0|2001:db8:1:0::/64. The xCAT object definition commands support to use nicnetworks.<nicname> as the sub attributes.
nictypes (nics.nictypes)
Comma-separated list of NIC types per NIC. <nic1>!<type1>,<nic2>!<type2>, e.g. eth0!Ethernet,ib0!Infiniband. The xCAT object definition commands support to use nictypes.<nicname> as the sub attributes.
nimserver (noderes.nimserver)
Not used for now. The NIM server for this node (as known by this node).
nodetype (nodetype.nodetype)
A comma-delimited list of characteristics of this node. Valid values: ppc, blade, vm (virtual machine), osi (OS image), mm, mn, rsa, switch.
ondiscover (chain.ondiscover)
This attribute is currently not used by xCAT. The “nodediscover” operation is always done during node discovery.
os (nodetype.os)
The operating system deployed on this node. Valid values: AIX, rhels*,rhelc*, rhas*,centos*,SL*, fedora*, sles* (where * is the version #). As a special case, if this is set to “boottarget”, then it will use the initrd/kernel/parameters specified in the row in the boottarget table in which boottarget.bprofile equals nodetype.profile.
osvolume (storage.osvolume)
Specification of what storage to place the node OS image onto. Examples include:
localdisk (Install to first non-FC attached disk) usbdisk (Install to first USB mass storage device seen) wwn=0x50000393c813840c (Install to storage device with given WWN)
otherinterfaces (hosts.otherinterfaces)
Other IP addresses to add for this node. Format: -<ext>:<ip>,<intfhostname>:<ip>,...
ou (domain.ou)
For an LDAP described machine account (i.e. Active Directory), the orginaztional unit to place the system. If not set, defaults to cn=Computers,dc=your,dc=domain
parent (ppc.parent)
For LPARs: the CEC; for FSPs: the CEC; for CEC: the frame (if one exists); for BPA: the frame; for frame: the building block number (which consists 1 or more service nodes and compute/storage nodes that are serviced by them - optional).
passwd.HMC (ppcdirect.password)
Password of the FSP/BPA(for ASMI) and CEC/Frame(for DFM). If not filled in, xCAT will look in the passwd table for key=fsp. If not in the passwd table, the default used is admin.
passwd.admin (ppcdirect.password)
Password of the FSP/BPA(for ASMI) and CEC/Frame(for DFM). If not filled in, xCAT will look in the passwd table for key=fsp. If not in the passwd table, the default used is admin.
passwd.celogin (ppcdirect.password)
Password of the FSP/BPA(for ASMI) and CEC/Frame(for DFM). If not filled in, xCAT will look in the passwd table for key=fsp. If not in the passwd table, the default used is admin.
passwd.general (ppcdirect.password)
Password of the FSP/BPA(for ASMI) and CEC/Frame(for DFM). If not filled in, xCAT will look in the passwd table for key=fsp. If not in the passwd table, the default used is admin.
passwd.hscroot (ppcdirect.password)
Password of the FSP/BPA(for ASMI) and CEC/Frame(for DFM). If not filled in, xCAT will look in the passwd table for key=fsp. If not in the passwd table, the default used is admin.
password (ppchcp.password, mpa.password, websrv.password)
Password of the HMC or IVM. If not filled in, xCAT will look in the passwd table for key=hmc or key=ivm. If not in the passwd table, the default used is abc123 for HMCs and padmin for IVMs.
or
Password to use to access the management module. If not specified, the key=blade row in the passwd table is used as the default.
or
Password to use to access the web service.
postbootscripts (postscripts.postbootscripts)
Comma separated list of scripts that should be run on this node after diskfull installation or diskless boot. Each script can take zero or more parameters. For example: “script1 p1 p2,script2,...”. On AIX these scripts are run during the processing of /etc/inittab. On Linux they are run at the init.d time. xCAT automatically adds the scripts in the xcatdefaults.postbootscripts attribute to run first in the list.
postscripts (postscripts.postscripts)
Comma separated list of scripts that should be run on this node after diskfull installation or diskless boot. Each script can take zero or more parameters. For example: “script1 p1 p2,script2,...”. xCAT automatically adds the postscripts from the xcatdefaults.postscripts attribute of the table to run first on the nodes after install or diskless boot. For installation of RedHat, CentOS, Fedora, the scripts will be run before the reboot. For installation of SLES, the scripts will be run after the reboot but before the init.d process. For diskless deployment, the scripts will be run at the init.d time, and xCAT will automatically add the list of scripts from the postbootscripts attribute to run after postscripts list. For installation of AIX, the scripts will run after the reboot and acts the same as the postbootscripts attribute. For AIX, use the postbootscripts attribute. Support will be added in the future for the postscripts attribute to run the scripts before the reboot in AIX.
power (nodehm.power)
The method to use to control the power of the node. If not set, the mgt attribute will be used. Valid values: ipmi, blade, hmc, ivm, fsp, kvm, esx, rhevm. If “ipmi”, xCAT will search for this node in the ipmi table for more info. If “blade”, xCAT will search for this node in the mp table. If “hmc”, “ivm”, or “fsp”, xCAT will search for this node in the ppc table.
pprofile (ppc.pprofile)
The LPAR profile that will be used the next time the LPAR is powered on with rpower. For DFM, the pprofile attribute should be set to blank
prescripts-begin (prescripts.begin)
The scripts to be run at the beginning of the nodeset(Linux),
nimnodeset(AIX) or mkdsklsnode(AIX) command. The format is: [action1:]s1,s2...[|action2:s3,s4,s5...] where: - action1 and action2 for Linux are the nodeset actions specified in the command. For AIX, action1 and action1 can be 'diskless' for mkdsklsnode command' and 'standalone for nimnodeset command. - s1 and s2 are the scripts to run for action1 in order. - s3, s4, and s5 are the scripts to run for actions2. If actions are omitted, the scripts apply to all actions. Examples: myscript1,myscript2 (all actions) diskless:myscript1,myscript2 (AIX) install:myscript1,myscript2|netboot:myscript3 (Linux) All the scripts should be copied to /install/prescripts directory. The following two environment variables will be passed to each script: NODES a coma separated list of node names that need to run the script for ACTION current nodeset action. If '#xCAT setting:MAX_INSTANCE=number' is specified in the script, the script will get invoked for each node in parallel, but no more than number of instances will be invoked at at a time. If it is not specified, the script will be invoked once for all the nodes.
prescripts-end (prescripts.end)
The scripts to be run at the end of the nodeset(Linux),
nimnodeset(AIX),or mkdsklsnode(AIX) command. The format is the same as the 'begin' column.
primarynic (noderes.primarynic)
This attribute will be deprecated. All the used network interface will be determined by installnic. The network adapter on the node that will be used for xCAT management, the primarynic can be set to the network adapter name or the mac address or the keyword “mac” which means that the network interface specified by the mac address in the mac table will be used. Default is eth0.
productkey (prodkey.key)
The product key relevant to the aforementioned node/group and product combination
profile (nodetype.profile)
The string to use to locate a kickstart or autoyast template to use for OS deployment of this node. If the provmethod attribute is set to an osimage name, that takes precedence, and profile need not be defined. Otherwise, the os, profile, and arch are used to search for the files in /install/custom first, and then in /opt/xcat/share/xcat.
provmethod (nodetype.provmethod)
The provisioning method for node deployment. The valid values are install, netboot, statelite or an os image name from the osimage table. If an image name is specified, the osimage definition stored in the osimage table and the linuximage table (for Linux) or nimimage table (for AIX) are used to locate the files for templates, pkglists, syncfiles, etc. On Linux, if install, netboot or statelite is specified, the os, profile, and arch are used to search for the files in /install/custom first, and then in /opt/xcat/share/xcat.
rack (nodepos.rack)
The frame the node is in.
room (nodepos.room)
The room where the node is located.
routenames (noderes.routenames)
A comma separated list of route names that refer to rows in the routes table. These are the routes that should be defined on this node when it is deployed.
serial (vpd.serial)
The serial number of the node.
serialflow (nodehm.serialflow)
The flow control value of the serial port for this node. For SOL this is typically ‘hard’.
serialport (nodehm.serialport)
The serial port for this node, in the linux numbering style (0=COM1/ttyS0, 1=COM2/ttyS1). For SOL on IBM blades, this is typically 1. For rackmount IBM servers, this is typically 0.
serialspeed (nodehm.serialspeed)
The speed of the serial port for this node. For SOL this is typically 19200.
servicenode (noderes.servicenode)
A comma separated list of node names (as known by the management node) that provides most services for this node. The first service node on the list that is accessible will be used. The 2nd node on the list is generally considered to be the backup service node for this node when running commands like snmove.
setupconserver (servicenode.conserver)
Do we set up Conserver on this service node? Valid values:yes or 1, no or 0. If yes, configures and starts conserver daemon. If no or 0, it does not change the current state of the service.
setupdhcp (servicenode.dhcpserver)
Do we set up DHCP on this service node? Not supported on AIX. Valid values:yes or 1, no or 0. If yes, runs makedhcp -n. If no or 0, it does not change the current state of the service.
setupftp (servicenode.ftpserver)
Do we set up a ftp server on this service node? Not supported on AIX Valid values:yes or 1, no or 0. If yes, configure and start vsftpd. (You must manually install vsftpd on the service nodes before this.) If no or 0, it does not change the current state of the service. xCAT is not using ftp for compute nodes provisioning or any other xCAT features, so this attribute can be set to 0 if the ftp service will not be used for other purposes
setupipforward (servicenode.ipforward)
Do we set up ip forwarding on this service node? Valid values:yes or 1, no or 0. If no or 0, it does not change the current state of the service.
setupldap (servicenode.ldapserver)
Do we set up ldap caching proxy on this service node? Not supported on AIX. Valid values:yes or 1, no or 0. If no or 0, it does not change the current state of the service.
setupnameserver (servicenode.nameserver)
Do we set up DNS on this service node? Valid values: 2, 1, no or 0. If 2, creates named.conf as dns slave, using the management node as dns master, and starts named. If 1, creates named.conf file with forwarding to the management node and starts named. If no or 0, it does not change the current state of the service.
setupnfs (servicenode.nfsserver)
Do we set up file services (HTTP,FTP,or NFS) on this service node? For AIX will only setup NFS, not HTTP or FTP. Valid values:yes or 1, no or 0.If no or 0, it does not change the current state of the service.
setupnim (servicenode.nimserver)
Not used. Do we set up a NIM server on this service node? Valid values:yes or 1, no or 0. If no or 0, it does not change the current state of the service.
setupntp (servicenode.ntpserver)
Not used. Use setupntp postscript to setup a ntp server on this service node? Valid values:yes or 1, no or 0. If no or 0, it does not change the current state of the service.
setupproxydhcp (servicenode.proxydhcp)
Do we set up proxydhcp service on this node? valid values: yes or 1, no or 0. If yes, the proxydhcp daemon will be enabled on this node.
setuptftp (servicenode.tftpserver)
Do we set up TFTP on this service node? Not supported on AIX. Valid values:yes or 1, no or 0. If yes, configures and starts atftp. If no or 0, it does not change the current state of the service.
sfp (ppc.sfp)
The Service Focal Point of this Frame. This is the name of the HMC that is responsible for collecting hardware service events for this frame and all of the CECs within this frame.
side (vpd.side)
<BPA>-<port> or <FSP>-<port>. The side information for the BPA/FSP. The side attribute refers to which BPA/FSP, A or B, which is determined by the slot value returned from lsslp command. It also lists the physical port within each BPA/FSP which is determined by the IP address order from the lsslp response. This information is used internally when communicating with the BPAs/FSPs
slot (nodepos.slot)
The slot number of the blade in the chassis. For PCM, a comma-separated list of slot numbers is stored
slotid (mp.id)
The slot number of this blade in the BladeCenter chassis.
slots (mpa.slots)
The number of available slots in the chassis. For PCM, this attribute is used to store the number of slots in the following format: <slot rows>,<slot columns>,<slot orientation> Where:
<slot rows> = number of rows of slots in chassis <slot columns> = number of columns of slots in chassis <slot orientation> = set to 0 if slots are vertical, and set to 1 if slots of horizontal
storagcontroller (storage.controller)
The management address to attach/detach new volumes.
In the scenario involving multiple controllers, this data must be passed as argument rather than by table value
storagetype (storage.type)
The plugin used to drive storage configuration (e.g. svc)
supernode (ppc.supernode)
Indicates the connectivity of this CEC in the HFI network. A comma separated list of 2 ids. The first one is the supernode number the CEC is part of. The second one is the logical location number (0-3) of this CEC within the supernode.
supportedarchs (nodetype.supportedarchs)
Comma delimited list of architectures this node can execute.
supportproxydhcp (noderes.proxydhcp)
To specify whether the node supports proxydhcp protocol. Valid values: yes or 1, no or 0. Default value is yes.
switch (switch.switch)
The switch hostname.
switchinterface (switch.interface)
The interface name from the node perspective. For example, eth0. For the primary nic, it can be empty, the word “primary” or “primary:ethx” where ethx is the interface name.
switchport (switch.port)
The port number in the switch that this node is connected to. On a simple 1U switch, an administrator can generally enter the number as printed next to the ports, and xCAT will understand switch representation differences. On stacked switches or switches with line cards, administrators should usually use the CLI representation (i.e. 2/0/1 or 5/8). One notable exception is stacked SMC 8848M switches, in which you must add 56 for the proceeding switch, then the port number. For example, port 3 on the second switch in an SMC8848M stack would be 59
switchvlan (switch.vlan)
The ID for the tagged vlan that is created on this port using mkvlan and chvlan commands.
termport (nodehm.termport)
The port number on the terminal server that this node is connected to.
termserver (nodehm.termserver)
The hostname of the terminal server.
tftpdir (noderes.tftpdir)
The directory that roots this nodes contents from a tftp and related perspective. Used for NAS offload by using different mountpoints.
tftpserver (noderes.tftpserver)
The TFTP server for this node (as known by this node). If not set, it defaults to networks.tftpserver.
unit (nodepos.u)
The vertical position of the node in the frame
urlpath (mpa.urlpath)
URL path for the Chassis web interface. The full URL is built as follows: <hostname>/<urlpath>
usercomment (nodegroup.comments)
Any user-written notes.
userid (zvm.userid)
The z/VM userID of this node.
username (ppchcp.username, mpa.username, websrv.username)
Userid of the HMC or IVM. If not filled in, xCAT will look in the passwd table for key=hmc or key=ivm. If not in the passwd table, the default used is hscroot for HMCs and padmin for IVMs.
or
Userid to use to access the management module.
or
Userid to use to access the web service.
vmbeacon (vm.beacon)
This flag is used by xCAT to track the state of the identify LED with respect to the VM.
vmbootorder (vm.bootorder)
Boot sequence (i.e. net,hd)
vmcfgstore (vm.cfgstore)
Optional location for persistant storage separate of emulated hard drives for virtualization solutions that require persistant store to place configuration data
vmcluster (vm.cluster)
Specify to the underlying virtualization infrastructure a cluster membership for the hypervisor.
vmcpus (vm.cpus)
Number of CPUs the node should see.
vmhost (vm.host)
The system that currently hosts the VM
vmmanager (vm.mgr)
The function manager for the virtual machine
vmmaster (vm.master)
The name of a master image, if any, this virtual machine is linked to. This is generally set by clonevm and indicates the deletion of a master that would invalidate the storage of this virtual machine
vmmemory (vm.memory)
Megabytes of memory the VM currently should be set to.
vmnicnicmodel (vm.nicmodel)
Model of NICs that will be provided to VMs (i.e. e1000, rtl8139, virtio, etc)
vmnics (vm.nics)
Network configuration parameters. Of the general form [physnet:]interface,.. Generally, interface describes the vlan entity (default for native, tagged for tagged, vl[number] for a specific vlan. physnet is a virtual switch name or port description that is used for some virtualization technologies to construct virtual switches. hypervisor.netmap can map names to hypervisor specific layouts, or the descriptions described there may be used directly here where possible.
vmothersetting (vm.othersettings)
This allows specifying a semicolon delimited list of key->value pairs to include in a vmx file of VMware. For partitioning on normal power machines, this option is used to specify the hugepage and/or bsr information, the value is like:’hugepage:1,bsr=2’.
vmphyslots (vm.physlots)
Specify the physical slots drc index that will assigned to the partition, the delimiter is ‘,’, and the drc index must started with ‘0x’. For more details, please reference to manpage of ‘lsvm’.
vmstorage (vm.storage)
A list of storage files or devices to be used. i.e. dir:///cluster/vm/<nodename> or nfs://<server>/path/to/folder/
vmstoragecache (vm.storagecache)
Select caching scheme to employ. E.g. KVM understands ‘none’, ‘writethrough’ and ‘writeback’
vmstorageformat (vm.storageformat)
Select disk format to use by default (e.g. raw versus qcow2)
vmstoragemodel (vm.storagemodel)
Model of storage devices to provide to guest
vmtextconsole (vm.textconsole)
Tracks the Psuedo-TTY that maps to the serial port or console of a VM
vmvirtflags (vm.virtflags)
General flags used by the virtualization method. For example, in Xen it could, among other things, specify paravirtualized setup, or direct kernel boot. For a hypervisor/dom0 entry, it is the virtualization method (i.e. “xen”). For KVM, the following flag=value pairs are recognized:
imageformat=[raw|fullraw|qcow2] raw is a generic sparse file that allocates storage on demand fullraw is a generic, non-sparse file that preallocates all space qcow2 is a sparse, copy-on-write capable format implemented at the virtualization layer rather than the filesystem level clonemethod=[qemu-img|reflink] qemu-img allows use of qcow2 to generate virtualization layer copy-on-write reflink uses a generic filesystem facility to clone the files on your behalf, but requires filesystem support such as btrfs placement_affinity=[migratable|user_migratable|pinned]
vmvncport (vm.vncport)
Tracks the current VNC display port (currently not meant to be set
webport (websrv.port)
The port of the web service.
wherevals (nodegroup.wherevals)
A list of “attr*val” pairs that can be used to determine the members of a dynamic group, the delimiter is ”::” and the operator * can be ==, =~, != or !~.
xcatmaster (noderes.xcatmaster)
The hostname of the xCAT service node (as known by this node). This acts as the default value for nfsserver and tftpserver, if they are not set. If xcatmaster is not set, the node will use whoever responds to its boot request as its master. For the directed bootp case for POWER, it will use the management node if xcatmaster is not set.
mkdef(1), chdef(1), lsdef(1), rmdef(1)
kit.7¶
kit - a logical object definition in the xCAT database.
kit Attributes: basename, description, isinternal, kitdeployparams, kitdir, kitname, ostype, release, version
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
basename (kit.basename)
The kit base name
description (kit.description)
The Kit description.
isinternal (kit.isinternal)
A flag to indicated if the Kit is internally used. When set to 1, the Kit is internal. If 0 or undefined, the kit is not internal.
kitdeployparams (kit.kitdeployparams)
The file containing the default deployment parameters for this Kit. These parameters are added to the OS Image definition.s list of deployment parameters when one or more Kit Components from this Kit are added to the OS Image.
kitdir (kit.kitdir)
The path to Kit Installation directory on the Mgt Node.
kitname (kit.kitname)
The unique generated kit name, when kit is added to the cluster.
ostype (kit.ostype)
The kit OS type. Linux or AIX.
release (kit.release)
The kit release
version (kit.version)
The kit version
mkdef(1), chdef(1), lsdef(1), rmdef(1)
kitcomponent.7¶
kitcomponent - a logical object definition in the xCAT database.
kitcomponent Attributes: basename, description, driverpacks, exlist, genimage_postinstall, kitcompdeps, kitcompname, kitname, kitpkgdeps, kitreponame, postbootscripts, prerequisite, release, serverroles, version
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
basename (kitcomponent.basename)
Kit Component basename.
description (kitcomponent.description)
The Kit component description.
driverpacks (kitcomponent.driverpacks)
Comma-separated List of driver package names. These must be full names like: pkg1-1.0-1.x86_64.rpm.
exlist (kitcomponent.exlist)
Exclude list file containing the files/directories to exclude when building a diskless image.
genimage_postinstall (kitcomponent.genimage_postinstall)
Comma-separated list of postinstall scripts that will run during the genimage.
kitcompdeps (kitcomponent.kitcompdeps)
Comma-separated list of kit components that this kit component depends on.
kitcompname (kitcomponent.kitcompname)
The unique Kit Component name. It is auto-generated when the parent Kit is added to the cluster.
kitname (kitcomponent.kitname)
The Kit name which this Kit Component belongs to.
kitpkgdeps (kitcomponent.kitpkgdeps)
Comma-separated list of packages that this kit component depends on.
kitreponame (kitcomponent.kitreponame)
The Kit Package Repository name which this Kit Component belongs to.
postbootscripts (kitcomponent.postbootscripts)
Comma-separated list of postbootscripts that will run during the node boot.
prerequisite (kitcomponent.prerequisite)
Prerequisite for this kit component, the prerequisite includes ospkgdeps,preinstall,preupgrade,preuninstall scripts
release (kitcomponent.release)
Kit Component release.
serverroles (kitcomponent.serverroles)
The types of servers that this Kit Component can install on. Valid types are: mgtnode, servicenode, compute
version (kitcomponent.version)
Kit Component version.
mkdef(1), chdef(1), lsdef(1), rmdef(1)
kitrepo.7¶
kitrepo - a logical object definition in the xCAT database.
kitrepo Attributes: compat_osbasenames, kitname, kitrepodir, kitreponame, osarch, osbasename, osmajorversion, osminorversion
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
compat_osbasenames (kitrepo.compat_osbasenames)
List of compatible OS base names.
kitname (kitrepo.kitname)
The Kit name which this Kit Package Repository belongs to.
kitrepodir (kitrepo.kitrepodir)
The path to Kit Repository directory on the Mgt Node.
kitreponame (kitrepo.kitreponame)
The unique generated kit repo package name, when kit is added to the cluster.
osarch (kitrepo.osarch)
The OS distro arch which this repository is based on.
osbasename (kitrepo.osbasename)
The OS distro name which this repository is based on.
osmajorversion (kitrepo.osmajorversion)
The OS distro major version which this repository is based on.
osminorversion (kitrepo.osminorversion)
The OS distro minor version which this repository is based on. If this attribute is not set, it means that this repo applies to all minor versions.
mkdef(1), chdef(1), lsdef(1), rmdef(1)
monitoring.7¶
monitoring - a logical object definition in the xCAT database.
monitoring Attributes: comments, disable, name, nodestatmon
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
comments (monitoring.comments)
Any user-written notes.
disable (monitoring.disable)
Set to ‘yes’ or ‘1’ to comment out this row.
name (monitoring.name)
The name of the mornitoring plug-in module. The plug-in must be put in /lib/perl/xCAT_monitoring/. See the man page for monstart for details.
nodestatmon (monitoring.nodestatmon)
Specifies if the monitoring plug-in is used to feed the node status to the xCAT cluster. Any one of the following values indicates “yes”: y, Y, yes, Yes, YES, 1. Any other value or blank (default), indicates “no”.
mkdef(1), chdef(1), lsdef(1), rmdef(1)
network.7¶
network - a logical object definition in the xCAT database.
network Attributes: ddnsdomain, dhcpserver, domain, dynamicrange, gateway, logservers, mask, mgtifname, nameservers, net, netname, nodehostname, ntpservers, staticrange, staticrangeincrement, tftpserver, usercomment, vlanid
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
ddnsdomain (networks.ddnsdomain)
A domain to be combined with nodename to construct FQDN for DDNS updates induced by DHCP. This is not passed down to the client as “domain”
dhcpserver (networks.dhcpserver)
The DHCP server that is servicing this network. Required to be explicitly set for pooled service node operation.
domain (networks.domain)
The DNS domain name (ex. cluster.com).
dynamicrange (networks.dynamicrange)
The IP address range used by DHCP to assign dynamic IP addresses for requests on this network. This should not overlap with entities expected to be configured with static host declarations, i.e. anything ever expected to be a node with an address registered in the mac table.
gateway (networks.gateway)
The network gateway. It can be set to an ip address or the keyword <xcatmaster>, the keyword <xcatmaster> indicates the cluster-facing ip address configured on this management node or service node. Leaving this field blank means that there is no gateway for this network.
logservers (networks.logservers)
The log servers for this network. Used in creating the DHCP network definition. Assumed to be the DHCP server if not set.
mask (networks.mask)
The network mask.
mgtifname (networks.mgtifname)
The interface name of the management/service node facing this network. !remote!<nicname> indicates a non-local network on a specific nic for relay DHCP.
nameservers (networks.nameservers)
A comma delimited list of DNS servers that each node in this network should use. This value will end up in the nameserver settings of the /etc/resolv.conf on each node in this network. If this attribute value is set to the IP address of an xCAT node, make sure DNS is running on it. In a hierarchical cluster, you can also set this attribute to “<xcatmaster>” to mean the DNS server for each node in this network should be the node that is managing it (either its service node or the management node). Used in creating the DHCP network definition, and DNS configuration.
net (networks.net)
The network address.
netname (networks.netname)
Name used to identify this network definition.
nodehostname (networks.nodehostname)
A regular expression used to specify node name to network-specific hostname. i.e. “/z/-secondary/” would mean that the hostname of “n1” would be n1-secondary on this network. By default, the nodename is assumed to equal the hostname, followed by nodename-interfacename.
ntpservers (networks.ntpservers)
The ntp servers for this network. Used in creating the DHCP network definition. Assumed to be the DHCP server if not set.
staticrange (networks.staticrange)
The IP address range used to dynamically assign static IPs to newly discovered nodes. This should not overlap with the dynamicrange nor overlap with entities that were manually assigned static IPs. The format for the attribute value is: <startip>-<endip>.
staticrangeincrement (networks.staticrangeincrement)
tftpserver (networks.tftpserver)
The TFTP server that is servicing this network. If not set, the DHCP server is assumed.
usercomment (networks.comments)
Any user-written notes.
vlanid (networks.vlanid)
The vlan ID if this network is within a vlan.
mkdef(1), chdef(1), lsdef(1), rmdef(1)
node.7¶
node - a logical object definition in the xCAT database.
node Attributes: addkcmdline, appstatus, appstatustime, arch, authdomain, bmc, bmcpassword, bmcport, bmcusername, cfgmgr, cfgmgtroles, cfgserver, chain, chassis, cmdmapping, cons, conserver, consoleondemand, cpucount, cputype, currchain, currstate, dhcpinterfaces, disksize, displayname, domainadminpassword, domainadminuser, domaintype, getmac, groups, hcp, height, hidden, hostcluster, hostinterface, hostmanager, hostnames, hosttype, hwtype, id, initrd, installnic, interface, ip, iscsipassword, iscsiserver, iscsitarget, iscsiuserid, kcmdline, kernel, mac, memory, mgt, micbridge, michost, micid, miconboot, micpowermgt, micvlog, migrationdest, monserver, mpa, mtm, nameservers, netboot, nfsdir, nfsserver, nicaliases, niccustomscripts, nicextraparams, nichostnameprefixes, nichostnamesuffixes, nicips, nicnetworks, nictypes, nimserver, node, nodetype, ondiscover, os, osvolume, otherinterfaces, ou, parent, passwd.HMC, passwd.admin, passwd.celogin, passwd.general, passwd.hscroot, password, postbootscripts, postscripts, power, pprofile, prescripts-begin, prescripts-end, primarynic, primarysn, productkey, profile, provmethod, rack, room, routenames, serial, serialflow, serialport, serialspeed, servicenode, setupconserver, setupdhcp, setupftp, setupipforward, setupldap, setupnameserver, setupnfs, setupnim, setupntp, setupproxydhcp, setuptftp, sfp, side, slot, slotid, slots, status, statustime, storagcontroller, storagetype, supernode, supportedarchs, supportproxydhcp, switch, switchinterface, switchport, switchvlan, termport, termserver, tftpdir, tftpserver, unit, updatestatus, updatestatustime, urlpath, usercomment, userid, username, vmbeacon, vmbootorder, vmcfgstore, vmcluster, vmcpus, vmhost, vmmanager, vmmaster, vmmemory, vmnicnicmodel, vmnics, vmothersetting, vmphyslots, vmstorage, vmstoragecache, vmstorageformat, vmstoragemodel, vmtextconsole, vmvirtflags, vmvncport, webport, xcatmaster, zonename
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
addkcmdline (bootparams.addkcmdline)
User specified one or more parameters to be passed to the kernel. For the kernel options need to be persistent after installation, specify them with prefix “R::”
appstatus (nodelist.appstatus)
A comma-delimited list of application status. For example: ‘sshd=up,ftp=down,ll=down’
appstatustime (nodelist.appstatustime)
The date and time when appstatus was updated.
arch (nodetype.arch)
The hardware architecture of this node. Valid values: x86_64, ppc64, x86, ia64.
authdomain (domain.authdomain)
If a node should participate in an AD domain or Kerberos realm distinct from domain indicated in site, this field can be used to specify that
bmc (ipmi.bmc)
The hostname of the BMC adapater.
bmcpassword (ipmi.password)
The BMC password. If not specified, the key=ipmi row in the passwd table is used as the default.
bmcport (ipmi.bmcport)
In systems with selectable shared/dedicated ethernet ports, this parameter can be used to specify the preferred port. 0 means use the shared port, 1 means dedicated, blank is to not assign. The following special cases exist for IBM System x servers: For x3755 M3 systems, 0 means use the dedicated port, 1 means shared, blank is to not assign. For certain systems which have a mezzaine or ML2 adapter, there is a second value to include: For x3750 M4 (Model 8722): 0 2 1st 1Gbps interface for LOM 0 0 1st 10Gbps interface for LOM 0 3 2nd 1Gbps interface for LOM 0 1 2nd 10Gbps interface for LOM For x3750 M4 (Model 8752), x3850/3950 X6, dx360 M4, x3550 M4, and x3650 M4: 0 Shared (1st onboard interface) 1 Dedicated 2 0 First interface on ML2 or mezzanine adapter 2 1 Second interface on ML2 or mezzanine adapter 2 2 Third interface on ML2 or mezzanine adapter 2 3 Fourth interface on ML2 or mezzanine adapter
bmcusername (ipmi.username)
The BMC userid. If not specified, the key=ipmi row in the passwd table is used as the default.
cfgmgr (cfgmgt.cfgmgr)
The name of the configuration manager service. Currently ‘chef’ and ‘puppet’ are supported services.
cfgmgtroles (cfgmgt.roles)
The roles associated with this node as recognized by the cfgmgr for the software that is to be installed and configured. These role names map to chef recipes or puppet manifest classes that should be used for this node. For example, chef OpenStack cookbooks have roles such as mysql-master,keystone, glance, nova-controller, nova-conductor, cinder-all.
cfgserver (cfgmgt.cfgserver)
The xCAT node name of the chef server or puppet master
chain (chain.chain)
A comma-delimited chain of actions to be performed automatically when this node is discovered. (“Discovered” means a node booted, but xCAT and DHCP did not recognize the MAC of this node. In this situation, xCAT initiates the discovery process, the last step of which is to run the operations listed in this chain attribute, one by one.) Valid values: boot or reboot, install or netboot, runcmd=<cmd>, runimage=<URL>, shell, standby. (Default - same as no chain - it will do only the discovery.). Example, for BMC machines use: runcmd=bmcsetup,shell.
chassis (nodepos.chassis)
The BladeCenter chassis the blade is in.
cmdmapping (nodehm.cmdmapping)
The fully qualified name of the file that stores the mapping between PCM hardware management commands and xCAT/third-party hardware management commands for a particular type of hardware device. Only used by PCM.
cons (nodehm.cons)
The console method. If nodehm.serialport is set, this will default to the nodehm.mgt setting, otherwise it defaults to unused. Valid values: cyclades, mrv, or the values valid for the mgt attribute.
conserver (nodehm.conserver)
The hostname of the machine where the conserver daemon is running. If not set, the default is the xCAT management node.
consoleondemand (nodehm.consoleondemand)
This overrides the value from site.consoleondemand; (0=no, 1=yes). Default is the result from site.consoleondemand.
cpucount (hwinv.cpucount)
The number of cpus for the node.
cputype (hwinv.cputype)
The cpu model name for the node.
currchain (chain.currchain)
The chain steps still left to do for this node. This attribute will be automatically adjusted by xCAT while xCAT-genesis is running on the node (either during node discovery or a special operation like firmware update). During node discovery, this attribute is initialized from the chain attribute and updated as the chain steps are executed.
currstate (chain.currstate)
The current or next chain step to be executed on this node by xCAT-genesis. Set by xCAT during node discovery or as a result of nodeset.
dhcpinterfaces (servicenode.dhcpinterfaces)
The network interfaces DHCP server should listen on for the target node. This attribute can be used for management node and service nodes. If defined, it will override the values defined in site.dhcpinterfaces. This is a comma separated list of device names. !remote! indicates a non-local network for relay DHCP. For example: !remote!,eth0,eth1
disksize (hwinv.disksize)
The size of the disks for the node in GB.
displayname (mpa.displayname)
Alternative name for BladeCenter chassis. Only used by PCM.
domainadminpassword (domain.adminpassword)
Allow a node specific indication of Administrative user password for the domain. Most will want to ignore this in favor of passwd table.
domainadminuser (domain.adminuser)
Allow a node specific indication of Administrative user. Most will want to just use passwd table to indicate this once rather than by node.
domaintype (domain.type)
Type, if any, of authentication domain to manipulate. The only recognized value at the moment is activedirectory.
getmac (nodehm.getmac)
The method to use to get MAC address of the node with the getmac command. If not set, the mgt attribute will be used. Valid values: same as values for mgmt attribute.
groups (nodelist.groups)
A comma-delimited list of groups this node is a member of. Group names are arbitrary, except all nodes should be part of the ‘all’ group. Internal group names are designated by using __<groupname>. For example, __Unmanaged, could be the internal name for a group of nodes that is not managed by xCAT. Admins should avoid using the __ characters when defining their groups.
hcp (ppc.hcp, zvm.hcp)
The hardware control point for this node (HMC, IVM, Frame or CEC). Do not need to set for BPAs and FSPs.
or
The hardware control point for this node.
height (nodepos.height)
The server height in U(s).
hidden (nodelist.hidden)
Used to hide fsp and bpa definitions, 1 means not show them when running lsdef and nodels
hostcluster (hypervisor.cluster)
Specify to the underlying virtualization infrastructure a cluster membership for the hypervisor.
hostinterface (hypervisor.interface)
The definition of interfaces for the hypervisor. The format is [networkname:interfacename:bootprotocol:IP:netmask:gateway] that split with | for each interface
hostmanager (hypervisor.mgr)
The virtualization specific manager of this hypervisor when applicable
hostnames (hosts.hostnames)
Hostname aliases added to /etc/hosts for this node. Comma or blank separated list.
hosttype (hypervisor.type)
The plugin associated with hypervisor specific commands such as revacuate
hwtype (ppc.nodetype, zvm.nodetype, mp.nodetype, mic.nodetype)
The hardware type of the node. Only can be one of fsp, bpa, cec, frame, ivm, hmc and lpar
or
The node type. Valid values: cec (Central Electronic Complex), lpar (logical partition), zvm (z/VM host operating system), and vm (virtual machine).
or
The hardware type for mp node. Valid values: mm,cmm, blade.
or
The hardware type of the mic node. Generally, it is mic.
id (ppc.id, mp.id)
For LPARs: the LPAR numeric id; for CECs: the cage number; for Frames: the frame number.
or
The slot number of this blade in the BladeCenter chassis.
initrd (bootparams.initrd)
The initial ramdisk image that network boot actions should use (could be a DOS floppy or hard drive image if using memdisk as kernel)
installnic (noderes.installnic)
The network adapter on the node that will be used for OS deployment, the installnic can be set to the network adapter name or the mac address or the keyword “mac” which means that the network interface specified by the mac address in the mac table will be used. If not set, primarynic will be used. If primarynic is not set too, the keyword “mac” will be used as default.
interface (mac.interface)
The adapter interface name that will be used to install and manage the node. E.g. eth0 (for linux) or en0 (for AIX).)
ip (hosts.ip)
The IP address of the node. This is only used in makehosts. The rest of xCAT uses system name resolution to resolve node names to IP addresses.
iscsipassword (iscsi.passwd)
The password for the iscsi server containing the boot device for this node.
iscsiserver (iscsi.server)
The server containing the iscsi boot device for this node.
iscsitarget (iscsi.target)
The iscsi disk used for the boot device for this node. Filled in by xCAT.
iscsiuserid (iscsi.userid)
The userid of the iscsi server containing the boot device for this node.
kcmdline (bootparams.kcmdline)
Arguments to be passed to the kernel
kernel (bootparams.kernel)
The kernel that network boot actions should currently acquire and use. Note this could be a chained boot loader such as memdisk or a non-linux boot loader
mac (mac.mac)
The mac address or addresses for which xCAT will manage static bindings for this node. This may be simply a mac address, which would be bound to the node name (such as “01:02:03:04:05:0E”). This may also be a “|” delimited string of “mac address!hostname” format (such as “01:02:03:04:05:0E!node5|01:02:03:05:0F!node6-eth1”).
memory (hwinv.memory)
The size of the memory for the node in MB.
mgt (nodehm.mgt)
The method to use to do general hardware management of the node. This attribute is used as the default if power or getmac is not set. Valid values: ipmi, blade, hmc, ivm, fsp, bpa, kvm, esx, rhevm. See the power attribute for more details.
micbridge (mic.bridge)
The virtual bridge on the host node which the mic connected to.
michost (mic.host)
The host node which the mic card installed on.
micid (mic.id)
The device id of the mic node.
miconboot (mic.onboot)
Set mic to autoboot when mpss start. Valid values: yes|no. Default is yes.
micpowermgt (mic.powermgt)
Set the Power Management for mic node. This attribute is used to set the power management state that mic may get into when it is idle. Four states can be set: cpufreq, corec6, pc3 and pc6. The valid value for powermgt attribute should be [cpufreq=<on|off>]![corec6=<on|off>]![pc3=<on|off>]![pc6=<on|off>]. e.g. cpufreq=on!corec6=off!pc3=on!pc6=off. Refer to the doc of mic to get more information for power management.
micvlog (mic.vlog)
Set the Verbose Log to console. Valid values: yes|no. Default is no.
migrationdest (vm.migrationdest)
A noderange representing candidate destinations for migration (i.e. similar systems, same SAN, or other criteria that xCAT can use
monserver (noderes.monserver)
The monitoring aggregation point for this node. The format is “x,y” where x is the ip address as known by the management node and y is the ip address as known by the node.
mpa (mp.mpa)
The managment module used to control this blade.
mtm (vpd.mtm)
The machine type and model number of the node. E.g. 7984-6BU
nameservers (noderes.nameservers)
An optional node/group specific override for name server list. Most people want to stick to site or network defined nameserver configuration.
netboot (noderes.netboot)
The type of network booting to use for this node. Valid values: pxe or xnba for x86* architecture, yaboot for POWER architecture, grub2-tftp and grub2-http for RHEL7 on Power and all the os deployment on Power LE. Notice: yaboot is not supported from rhels7 on Power,use grub2-tftp or grub2-http instead, the difference between the 2 is the file transfer protocol(i.e, http or tftp)
nfsdir (noderes.nfsdir)
The path that should be mounted from the NFS server.
nfsserver (noderes.nfsserver)
The NFS or HTTP server for this node (as known by this node).
nicaliases (nics.nicaliases)
Comma-separated list of hostname aliases for each NIC.
Format: eth0!<alias list>,eth1!<alias1 list>|<alias2 list> For multiple aliases per nic use a space-separated list. For example: eth0!moe larry curly,eth1!tom|jerry
niccustomscripts (nics.niccustomscripts)
Comma-separated list of custom scripts per NIC. <nic1>!<script1>,<nic2>!<script2>, e.g. eth0!configeth eth0, ib0!configib ib0. The xCAT object definition commands support to use niccustomscripts.<nicname> as the sub attribute
.
nicextraparams (nics.nicextraparams)
Comma-separated list of extra parameters that will be used for each NIC configuration.
If only one ip address is associated with each NIC: <nic1>!<param1=value1 param2=value2>,<nic2>!<param3=value3>, for example, eth0!MTU=1500,ib0!MTU=65520 CONNECTED_MODE=yes. If multiple ip addresses are associated with each NIC: <nic1>!<param1=value1 param2=value2>|<param3=value3>,<nic2>!<param4=value4 param5=value5>|<param6=value6>, for example, eth0!MTU=1500|MTU=1460,ib0!MTU=65520 CONNECTED_MODE=yes. The xCAT object definition commands support to use nicextraparams.<nicname> as the sub attributes.
nichostnameprefixes (nics.nichostnameprefixes)
Comma-separated list of hostname prefixes per NIC.
If only one ip address is associated with each NIC: <nic1>!<ext1>,<nic2>!<ext2>,..., for example, eth0!eth0-,ib0!ib- If multiple ip addresses are associcated with each NIC: <nic1>!<ext1>|<ext2>,<nic2>!<ext1>|<ext2>,..., for example, eth0!eth0-|eth0-ipv6i-,ib0!ib-|ib-ipv6-. The xCAT object definition commands support to use nichostnameprefixes.<nicname> as the sub attributes. Note: According to DNS rules a hostname must be a text string up to 24 characters drawn from the alphabet (A-Z), digits (0-9), minus sign (-),and period (.). When you are specifying "nichostnameprefixes" or "nicaliases" make sure the resulting hostnames will conform to this naming convention
nichostnamesuffixes (nics.nichostnamesuffixes)
Comma-separated list of hostname suffixes per NIC.
If only one ip address is associated with each NIC: <nic1>!<ext1>,<nic2>!<ext2>,..., for example, eth0!-eth0,ib0!-ib0 If multiple ip addresses are associcated with each NIC: <nic1>!<ext1>|<ext2>,<nic2>!<ext1>|<ext2>,..., for example, eth0!-eth0|-eth0-ipv6,ib0!-ib0|-ib0-ipv6. The xCAT object definition commands support to use nichostnamesuffixes.<nicname> as the sub attributes. Note: According to DNS rules a hostname must be a text string up to 24 characters drawn from the alphabet (A-Z), digits (0-9), minus sign (-),and period (.). When you are specifying "nichostnamesuffixes" or "nicaliases" make sure the resulting hostnames will conform to this naming convention
nicips (nics.nicips)
Comma-separated list of IP addresses per NIC. To specify one ip address per NIC:
<nic1>!<ip1>,<nic2>!<ip2>,..., for example, eth0!10.0.0.100,ib0!11.0.0.100 To specify multiple ip addresses per NIC: <nic1>!<ip1>|<ip2>,<nic2>!<ip1>|<ip2>,..., for example, eth0!10.0.0.100|fd55::214:5eff:fe15:849b,ib0!11.0.0.100|2001::214:5eff:fe15:849a. The xCAT object definition commands support to use nicips.<nicname> as the sub attributes. Note: The primary IP address must also be stored in the hosts.ip attribute. The nichostnamesuffixes should specify one hostname suffix for each ip address.
nicnetworks (nics.nicnetworks)
Comma-separated list of networks connected to each NIC.
If only one ip address is associated with each NIC: <nic1>!<network1>,<nic2>!<network2>, for example, eth0!10_0_0_0-255_255_0_0, ib0!11_0_0_0-255_255_0_0 If multiple ip addresses are associated with each NIC: <nic1>!<network1>|<network2>,<nic2>!<network1>|<network2>, for example, eth0!10_0_0_0-255_255_0_0|fd55:faaf:e1ab:336::/64,ib0!11_0_0_0-255_255_0_0|2001:db8:1:0::/64. The xCAT object definition commands support to use nicnetworks.<nicname> as the sub attributes.
nictypes (nics.nictypes)
Comma-separated list of NIC types per NIC. <nic1>!<type1>,<nic2>!<type2>, e.g. eth0!Ethernet,ib0!Infiniband. The xCAT object definition commands support to use nictypes.<nicname> as the sub attributes.
nimserver (noderes.nimserver)
Not used for now. The NIM server for this node (as known by this node).
node (nodelist.node)
The hostname of a node in the cluster.
nodetype (nodetype.nodetype)
A comma-delimited list of characteristics of this node. Valid values: ppc, blade, vm (virtual machine), osi (OS image), mm, mn, rsa, switch.
ondiscover (chain.ondiscover)
This attribute is currently not used by xCAT. The “nodediscover” operation is always done during node discovery.
os (nodetype.os)
The operating system deployed on this node. Valid values: AIX, rhels*,rhelc*, rhas*,centos*,SL*, fedora*, sles* (where * is the version #). As a special case, if this is set to “boottarget”, then it will use the initrd/kernel/parameters specified in the row in the boottarget table in which boottarget.bprofile equals nodetype.profile.
osvolume (storage.osvolume)
Specification of what storage to place the node OS image onto. Examples include:
localdisk (Install to first non-FC attached disk) usbdisk (Install to first USB mass storage device seen) wwn=0x50000393c813840c (Install to storage device with given WWN)
otherinterfaces (hosts.otherinterfaces)
Other IP addresses to add for this node. Format: -<ext>:<ip>,<intfhostname>:<ip>,...
ou (domain.ou)
For an LDAP described machine account (i.e. Active Directory), the orginaztional unit to place the system. If not set, defaults to cn=Computers,dc=your,dc=domain
parent (ppc.parent)
For LPARs: the CEC; for FSPs: the CEC; for CEC: the frame (if one exists); for BPA: the frame; for frame: the building block number (which consists 1 or more service nodes and compute/storage nodes that are serviced by them - optional).
passwd.HMC (ppcdirect.password)
Password of the FSP/BPA(for ASMI) and CEC/Frame(for DFM). If not filled in, xCAT will look in the passwd table for key=fsp. If not in the passwd table, the default used is admin.
passwd.admin (ppcdirect.password)
Password of the FSP/BPA(for ASMI) and CEC/Frame(for DFM). If not filled in, xCAT will look in the passwd table for key=fsp. If not in the passwd table, the default used is admin.
passwd.celogin (ppcdirect.password)
Password of the FSP/BPA(for ASMI) and CEC/Frame(for DFM). If not filled in, xCAT will look in the passwd table for key=fsp. If not in the passwd table, the default used is admin.
passwd.general (ppcdirect.password)
Password of the FSP/BPA(for ASMI) and CEC/Frame(for DFM). If not filled in, xCAT will look in the passwd table for key=fsp. If not in the passwd table, the default used is admin.
passwd.hscroot (ppcdirect.password)
Password of the FSP/BPA(for ASMI) and CEC/Frame(for DFM). If not filled in, xCAT will look in the passwd table for key=fsp. If not in the passwd table, the default used is admin.
password (ppchcp.password, mpa.password, websrv.password)
Password of the HMC or IVM. If not filled in, xCAT will look in the passwd table for key=hmc or key=ivm. If not in the passwd table, the default used is abc123 for HMCs and padmin for IVMs.
or
Password to use to access the management module. If not specified, the key=blade row in the passwd table is used as the default.
or
Password to use to access the web service.
postbootscripts (postscripts.postbootscripts)
Comma separated list of scripts that should be run on this node after diskfull installation or diskless boot. Each script can take zero or more parameters. For example: “script1 p1 p2,script2,...”. On AIX these scripts are run during the processing of /etc/inittab. On Linux they are run at the init.d time. xCAT automatically adds the scripts in the xcatdefaults.postbootscripts attribute to run first in the list.
postscripts (postscripts.postscripts)
Comma separated list of scripts that should be run on this node after diskfull installation or diskless boot. Each script can take zero or more parameters. For example: “script1 p1 p2,script2,...”. xCAT automatically adds the postscripts from the xcatdefaults.postscripts attribute of the table to run first on the nodes after install or diskless boot. For installation of RedHat, CentOS, Fedora, the scripts will be run before the reboot. For installation of SLES, the scripts will be run after the reboot but before the init.d process. For diskless deployment, the scripts will be run at the init.d time, and xCAT will automatically add the list of scripts from the postbootscripts attribute to run after postscripts list. For installation of AIX, the scripts will run after the reboot and acts the same as the postbootscripts attribute. For AIX, use the postbootscripts attribute. Support will be added in the future for the postscripts attribute to run the scripts before the reboot in AIX.
power (nodehm.power)
The method to use to control the power of the node. If not set, the mgt attribute will be used. Valid values: ipmi, blade, hmc, ivm, fsp, kvm, esx, rhevm. If “ipmi”, xCAT will search for this node in the ipmi table for more info. If “blade”, xCAT will search for this node in the mp table. If “hmc”, “ivm”, or “fsp”, xCAT will search for this node in the ppc table.
pprofile (ppc.pprofile)
The LPAR profile that will be used the next time the LPAR is powered on with rpower. For DFM, the pprofile attribute should be set to blank
prescripts-begin (prescripts.begin)
The scripts to be run at the beginning of the nodeset(Linux),
nimnodeset(AIX) or mkdsklsnode(AIX) command. The format is: [action1:]s1,s2...[|action2:s3,s4,s5...] where: - action1 and action2 for Linux are the nodeset actions specified in the command. For AIX, action1 and action1 can be 'diskless' for mkdsklsnode command' and 'standalone for nimnodeset command. - s1 and s2 are the scripts to run for action1 in order. - s3, s4, and s5 are the scripts to run for actions2. If actions are omitted, the scripts apply to all actions. Examples: myscript1,myscript2 (all actions) diskless:myscript1,myscript2 (AIX) install:myscript1,myscript2|netboot:myscript3 (Linux) All the scripts should be copied to /install/prescripts directory. The following two environment variables will be passed to each script: NODES a coma separated list of node names that need to run the script for ACTION current nodeset action. If '#xCAT setting:MAX_INSTANCE=number' is specified in the script, the script will get invoked for each node in parallel, but no more than number of instances will be invoked at at a time. If it is not specified, the script will be invoked once for all the nodes.
prescripts-end (prescripts.end)
The scripts to be run at the end of the nodeset(Linux),
nimnodeset(AIX),or mkdsklsnode(AIX) command. The format is the same as the 'begin' column.
primarynic (noderes.primarynic)
This attribute will be deprecated. All the used network interface will be determined by installnic. The network adapter on the node that will be used for xCAT management, the primarynic can be set to the network adapter name or the mac address or the keyword “mac” which means that the network interface specified by the mac address in the mac table will be used. Default is eth0.
primarysn (nodelist.primarysn)
Not used currently. The primary servicenode, used by this node.
productkey (prodkey.key)
The product key relevant to the aforementioned node/group and product combination
profile (nodetype.profile)
The string to use to locate a kickstart or autoyast template to use for OS deployment of this node. If the provmethod attribute is set to an osimage name, that takes precedence, and profile need not be defined. Otherwise, the os, profile, and arch are used to search for the files in /install/custom first, and then in /opt/xcat/share/xcat.
provmethod (nodetype.provmethod)
The provisioning method for node deployment. The valid values are install, netboot, statelite or an os image name from the osimage table. If an image name is specified, the osimage definition stored in the osimage table and the linuximage table (for Linux) or nimimage table (for AIX) are used to locate the files for templates, pkglists, syncfiles, etc. On Linux, if install, netboot or statelite is specified, the os, profile, and arch are used to search for the files in /install/custom first, and then in /opt/xcat/share/xcat.
rack (nodepos.rack)
The frame the node is in.
room (nodepos.room)
The room where the node is located.
routenames (noderes.routenames)
A comma separated list of route names that refer to rows in the routes table. These are the routes that should be defined on this node when it is deployed.
serial (vpd.serial)
The serial number of the node.
serialflow (nodehm.serialflow)
The flow control value of the serial port for this node. For SOL this is typically ‘hard’.
serialport (nodehm.serialport)
The serial port for this node, in the linux numbering style (0=COM1/ttyS0, 1=COM2/ttyS1). For SOL on IBM blades, this is typically 1. For rackmount IBM servers, this is typically 0.
serialspeed (nodehm.serialspeed)
The speed of the serial port for this node. For SOL this is typically 19200.
servicenode (noderes.servicenode)
A comma separated list of node names (as known by the management node) that provides most services for this node. The first service node on the list that is accessible will be used. The 2nd node on the list is generally considered to be the backup service node for this node when running commands like snmove.
setupconserver (servicenode.conserver)
Do we set up Conserver on this service node? Valid values:yes or 1, no or 0. If yes, configures and starts conserver daemon. If no or 0, it does not change the current state of the service.
setupdhcp (servicenode.dhcpserver)
Do we set up DHCP on this service node? Not supported on AIX. Valid values:yes or 1, no or 0. If yes, runs makedhcp -n. If no or 0, it does not change the current state of the service.
setupftp (servicenode.ftpserver)
Do we set up a ftp server on this service node? Not supported on AIX Valid values:yes or 1, no or 0. If yes, configure and start vsftpd. (You must manually install vsftpd on the service nodes before this.) If no or 0, it does not change the current state of the service. xCAT is not using ftp for compute nodes provisioning or any other xCAT features, so this attribute can be set to 0 if the ftp service will not be used for other purposes
setupipforward (servicenode.ipforward)
Do we set up ip forwarding on this service node? Valid values:yes or 1, no or 0. If no or 0, it does not change the current state of the service.
setupldap (servicenode.ldapserver)
Do we set up ldap caching proxy on this service node? Not supported on AIX. Valid values:yes or 1, no or 0. If no or 0, it does not change the current state of the service.
setupnameserver (servicenode.nameserver)
Do we set up DNS on this service node? Valid values: 2, 1, no or 0. If 2, creates named.conf as dns slave, using the management node as dns master, and starts named. If 1, creates named.conf file with forwarding to the management node and starts named. If no or 0, it does not change the current state of the service.
setupnfs (servicenode.nfsserver)
Do we set up file services (HTTP,FTP,or NFS) on this service node? For AIX will only setup NFS, not HTTP or FTP. Valid values:yes or 1, no or 0.If no or 0, it does not change the current state of the service.
setupnim (servicenode.nimserver)
Not used. Do we set up a NIM server on this service node? Valid values:yes or 1, no or 0. If no or 0, it does not change the current state of the service.
setupntp (servicenode.ntpserver)
Not used. Use setupntp postscript to setup a ntp server on this service node? Valid values:yes or 1, no or 0. If no or 0, it does not change the current state of the service.
setupproxydhcp (servicenode.proxydhcp)
Do we set up proxydhcp service on this node? valid values: yes or 1, no or 0. If yes, the proxydhcp daemon will be enabled on this node.
setuptftp (servicenode.tftpserver)
Do we set up TFTP on this service node? Not supported on AIX. Valid values:yes or 1, no or 0. If yes, configures and starts atftp. If no or 0, it does not change the current state of the service.
sfp (ppc.sfp)
The Service Focal Point of this Frame. This is the name of the HMC that is responsible for collecting hardware service events for this frame and all of the CECs within this frame.
side (vpd.side)
<BPA>-<port> or <FSP>-<port>. The side information for the BPA/FSP. The side attribute refers to which BPA/FSP, A or B, which is determined by the slot value returned from lsslp command. It also lists the physical port within each BPA/FSP which is determined by the IP address order from the lsslp response. This information is used internally when communicating with the BPAs/FSPs
slot (nodepos.slot)
The slot number of the blade in the chassis. For PCM, a comma-separated list of slot numbers is stored
slotid (mp.id)
The slot number of this blade in the BladeCenter chassis.
slots (mpa.slots)
The number of available slots in the chassis. For PCM, this attribute is used to store the number of slots in the following format: <slot rows>,<slot columns>,<slot orientation> Where:
<slot rows> = number of rows of slots in chassis <slot columns> = number of columns of slots in chassis <slot orientation> = set to 0 if slots are vertical, and set to 1 if slots of horizontal
status (nodelist.status)
The current status of this node. This attribute will be set by xCAT software. Valid values: defined, booting, netbooting, booted, discovering, configuring, installing, alive, standingby, powering-off, unreachable. If blank, defined is assumed. The possible status change sequenses are: For installaton: defined->[discovering]->[configuring]->[standingby]->installing->booting->booted->[alive], For diskless deployment: defined->[discovering]->[configuring]->[standingby]->netbooting->booted->[alive], For booting: [alive/unreachable]->booting->[alive], For powering off: [alive]->powering-off->[unreachable], For monitoring: alive->unreachable. Discovering and configuring are for x Series dicovery process. Alive and unreachable are set only when there is a monitoring plug-in start monitor the node status for xCAT. Please note that the status values will not reflect the real node status if you change the state of the node from outside of xCAT (i.e. power off the node using HMC GUI).
statustime (nodelist.statustime)
The data and time when the status was updated.
storagcontroller (storage.controller)
The management address to attach/detach new volumes.
In the scenario involving multiple controllers, this data must be passed as argument rather than by table value
storagetype (storage.type)
The plugin used to drive storage configuration (e.g. svc)
supernode (ppc.supernode)
Indicates the connectivity of this CEC in the HFI network. A comma separated list of 2 ids. The first one is the supernode number the CEC is part of. The second one is the logical location number (0-3) of this CEC within the supernode.
supportedarchs (nodetype.supportedarchs)
Comma delimited list of architectures this node can execute.
supportproxydhcp (noderes.proxydhcp)
To specify whether the node supports proxydhcp protocol. Valid values: yes or 1, no or 0. Default value is yes.
switch (switch.switch)
The switch hostname.
switchinterface (switch.interface)
The interface name from the node perspective. For example, eth0. For the primary nic, it can be empty, the word “primary” or “primary:ethx” where ethx is the interface name.
switchport (switch.port)
The port number in the switch that this node is connected to. On a simple 1U switch, an administrator can generally enter the number as printed next to the ports, and xCAT will understand switch representation differences. On stacked switches or switches with line cards, administrators should usually use the CLI representation (i.e. 2/0/1 or 5/8). One notable exception is stacked SMC 8848M switches, in which you must add 56 for the proceeding switch, then the port number. For example, port 3 on the second switch in an SMC8848M stack would be 59
switchvlan (switch.vlan)
The ID for the tagged vlan that is created on this port using mkvlan and chvlan commands.
termport (nodehm.termport)
The port number on the terminal server that this node is connected to.
termserver (nodehm.termserver)
The hostname of the terminal server.
tftpdir (noderes.tftpdir)
The directory that roots this nodes contents from a tftp and related perspective. Used for NAS offload by using different mountpoints.
tftpserver (noderes.tftpserver)
The TFTP server for this node (as known by this node). If not set, it defaults to networks.tftpserver.
unit (nodepos.u)
The vertical position of the node in the frame
updatestatus (nodelist.updatestatus)
The current node update status. Valid states are synced, out-of-sync,syncing,failed.
updatestatustime (nodelist.updatestatustime)
The date and time when the updatestatus was updated.
urlpath (mpa.urlpath)
URL path for the Chassis web interface. The full URL is built as follows: <hostname>/<urlpath>
usercomment (nodelist.comments)
Any user-written notes.
userid (zvm.userid)
The z/VM userID of this node.
username (ppchcp.username, mpa.username, websrv.username)
Userid of the HMC or IVM. If not filled in, xCAT will look in the passwd table for key=hmc or key=ivm. If not in the passwd table, the default used is hscroot for HMCs and padmin for IVMs.
or
Userid to use to access the management module.
or
Userid to use to access the web service.
vmbeacon (vm.beacon)
This flag is used by xCAT to track the state of the identify LED with respect to the VM.
vmbootorder (vm.bootorder)
Boot sequence (i.e. net,hd)
vmcfgstore (vm.cfgstore)
Optional location for persistant storage separate of emulated hard drives for virtualization solutions that require persistant store to place configuration data
vmcluster (vm.cluster)
Specify to the underlying virtualization infrastructure a cluster membership for the hypervisor.
vmcpus (vm.cpus)
Number of CPUs the node should see.
vmhost (vm.host)
The system that currently hosts the VM
vmmanager (vm.mgr)
The function manager for the virtual machine
vmmaster (vm.master)
The name of a master image, if any, this virtual machine is linked to. This is generally set by clonevm and indicates the deletion of a master that would invalidate the storage of this virtual machine
vmmemory (vm.memory)
Megabytes of memory the VM currently should be set to.
vmnicnicmodel (vm.nicmodel)
Model of NICs that will be provided to VMs (i.e. e1000, rtl8139, virtio, etc)
vmnics (vm.nics)
Network configuration parameters. Of the general form [physnet:]interface,.. Generally, interface describes the vlan entity (default for native, tagged for tagged, vl[number] for a specific vlan. physnet is a virtual switch name or port description that is used for some virtualization technologies to construct virtual switches. hypervisor.netmap can map names to hypervisor specific layouts, or the descriptions described there may be used directly here where possible.
vmothersetting (vm.othersettings)
This allows specifying a semicolon delimited list of key->value pairs to include in a vmx file of VMware. For partitioning on normal power machines, this option is used to specify the hugepage and/or bsr information, the value is like:’hugepage:1,bsr=2’.
vmphyslots (vm.physlots)
Specify the physical slots drc index that will assigned to the partition, the delimiter is ‘,’, and the drc index must started with ‘0x’. For more details, please reference to manpage of ‘lsvm’.
vmstorage (vm.storage)
A list of storage files or devices to be used. i.e. dir:///cluster/vm/<nodename> or nfs://<server>/path/to/folder/
vmstoragecache (vm.storagecache)
Select caching scheme to employ. E.g. KVM understands ‘none’, ‘writethrough’ and ‘writeback’
vmstorageformat (vm.storageformat)
Select disk format to use by default (e.g. raw versus qcow2)
vmstoragemodel (vm.storagemodel)
Model of storage devices to provide to guest
vmtextconsole (vm.textconsole)
Tracks the Psuedo-TTY that maps to the serial port or console of a VM
vmvirtflags (vm.virtflags)
General flags used by the virtualization method. For example, in Xen it could, among other things, specify paravirtualized setup, or direct kernel boot. For a hypervisor/dom0 entry, it is the virtualization method (i.e. “xen”). For KVM, the following flag=value pairs are recognized:
imageformat=[raw|fullraw|qcow2] raw is a generic sparse file that allocates storage on demand fullraw is a generic, non-sparse file that preallocates all space qcow2 is a sparse, copy-on-write capable format implemented at the virtualization layer rather than the filesystem level clonemethod=[qemu-img|reflink] qemu-img allows use of qcow2 to generate virtualization layer copy-on-write reflink uses a generic filesystem facility to clone the files on your behalf, but requires filesystem support such as btrfs placement_affinity=[migratable|user_migratable|pinned]
vmvncport (vm.vncport)
Tracks the current VNC display port (currently not meant to be set
webport (websrv.port)
The port of the web service.
xcatmaster (noderes.xcatmaster)
The hostname of the xCAT service node (as known by this node). This acts as the default value for nfsserver and tftpserver, if they are not set. If xcatmaster is not set, the node will use whoever responds to its boot request as its master. For the directed bootp case for POWER, it will use the management node if xcatmaster is not set.
zonename (nodelist.zonename)
The name of the zone to which the node is currently assigned. If undefined, then it is not assigned to any zone.
mkdef(1), chdef(1), lsdef(1), rmdef(1)
notification.7¶
notification - a logical object definition in the xCAT database.
notification Attributes: comments, filename, tableops, tables
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
comments (notification.comments)
Any user-written notes.
filename (notification.filename)
The path name of a file that implements the callback routine when the monitored table changes. Can be a perl module or a command. See the regnotif man page for details.
tableops (notification.tableops)
Specifies the table operation to monitor for. Valid values: “d” (rows deleted), “a” (rows added), “u” (rows updated).
tables (notification.tables)
Comma-separated list of xCAT database tables to monitor.
mkdef(1), chdef(1), lsdef(1), rmdef(1)
osdistro.7¶
osdistro - a logical object definition in the xCAT database.
osdistro Attributes: arch, basename, dirpaths, majorversion, minorversion, osdistroname, type
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
arch (osdistro.arch)
The OS distro arch (e.g. x86_64)
basename (osdistro.basename)
The OS base name (e.g. rhels)
dirpaths (osdistro.dirpaths)
Directory paths where OS distro is store. There could be multiple paths if OS distro has more than one ISO image. (e.g. /install/rhels6.2/x86_64,...)
majorversion (osdistro.majorversion)
The OS distro major version.(e.g. 6)
minorversion (osdistro.minorversion)
The OS distro minor version. (e.g. 2)
osdistroname (osdistro.osdistroname)
Unique name (e.g. rhels6.2-x86_64)
type (osdistro.type)
Linux or AIX
mkdef(1), chdef(1), lsdef(1), rmdef(1)
osdistroupdate.7¶
osdistroupdate - a logical object definition in the xCAT database.
osdistroupdate Attributes: dirpath, downloadtime, osdistroname, osupdatename, usercomment
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
dirpath (osdistroupdate.dirpath)
Path to where OS distro update is stored. (e.g. /install/osdistroupdates/rhels6.2-x86_64-20120716-update)
downloadtime (osdistroupdate.downloadtime)
The timestamp when OS distro update was downloaded..
osdistroname (osdistroupdate.osdistroname)
The OS distro name to update. (e.g. rhels)
osupdatename (osdistroupdate.osupdatename)
Name of OS update. (e.g. rhn-update1)
usercomment (osdistroupdate.comments)
Any user-written notes.
mkdef(1), chdef(1), lsdef(1), rmdef(1)
osimage.7¶
osimage - a logical object definition in the xCAT database.
osimage Attributes: addkcmdline, boottarget, bosinst_data, cfmdir, configdump, crashkernelsize, description, driverupdatesrc, dump, exlist, fb_script, groups, home, image_data, imagename, imagetype, installp_bundle, installto, isdeletable, kerneldir, kernelver, kitcomponents, krpmver, lpp_source, mksysb, netdrivers, nimmethod, nimtype, nodebootif, osarch, osdistroname, osname, osupdatename, osvers, otherifce, otherpkgdir, otherpkglist, otherpkgs, paging, partitionfile, permission, pkgdir, pkglist, postbootscripts, postinstall, postscripts, profile, provmethod, resolv_conf, root, rootfstype, rootimgdir, script, serverrole, shared_home, shared_root, spot, synclists, template, tmp, usercomment, winpepath
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
addkcmdline (linuximage.addkcmdline)
User specified arguments to be passed to the kernel. The user arguments are appended to xCAT.s default kernel arguments. For the kernel options need to be persistent after installation, specify them with prefix “R::”. This attribute is ignored if linuximage.boottarget is set.
boottarget (linuximage.boottarget)
The name of the boottarget definition. When this attribute is set, xCAT will use the kernel, initrd and kernel params defined in the boottarget definition instead of the default.
bosinst_data (nimimage.bosinst_data)
The name of a NIM bosinst_data resource.
cfmdir (osimage.cfmdir)
CFM directory name for PCM. Set to /install/osimages/<osimage name>/cfmdir by PCM.
configdump (nimimage.configdump)
Specifies the type of system dump to be collected. The values are selective, full, and none. The default is selective.
crashkernelsize (linuximage.crashkernelsize)
the size that assigned to the kdump kernel. If the kernel size is not set, 256M will be the default value.
description (osimage.description)
OS Image Description
driverupdatesrc (linuximage.driverupdatesrc)
The source of the drivers which need to be loaded during the boot. Two types of driver update source are supported: Driver update disk and Driver rpm package. The value for this attribute should be comma separated sources. Each source should be the format tab:full_path_of_srouce_file. The tab keyword can be: dud (for Driver update disk) and rpm (for driver rpm). If missing the tab, the rpm format is the default. e.g. dud:/install/dud/dd.img,rpm:/install/rpm/d.rpm
dump (linuximage.dump, nimimage.dump)
The NFS directory to hold the Linux kernel dump file (vmcore) when the node with this image crashes, its format is “nfs://<nfs_server_ip>/<kdump_path>”. If you want to use the node’s “xcatmaster” (its SN or MN), <nfs_server_ip> can be left blank. For example, “nfs:///<kdump_path>” means the NFS directory to hold the kernel dump file is on the node’s SN, or MN if there’s no SN.
or
The name of the NIM dump resource.
exlist (linuximage.exlist)
The fully qualified name of the file that stores the file names and directory names that will be excluded from the image during packimage command. It is used for diskless image only.
fb_script (nimimage.fb_script)
The name of a NIM fb_script resource.
groups (osimage.groups)
A comma-delimited list of image groups of which this image is a member. Image groups can be used in the litefile and litetree table instead of a single image name. Group names are arbitrary.
home (nimimage.home)
The name of the NIM home resource.
image_data (nimimage.image_data)
The name of a NIM image_data resource.
imagename (osimage.imagename)
The name of this xCAT OS image definition.
imagetype (osimage.imagetype)
The type of operating system image this definition represents (linux,AIX).
installp_bundle (nimimage.installp_bundle)
One or more comma separated NIM installp_bundle resources.
installto (winimage.installto)
The disk and partition that the Windows will be deployed to. The valid format is <disk>:<partition>. If not set, default value is 0:1 for bios boot mode(legacy) and 0:3 for uefi boot mode; If setting to 1, it means 1:1 for bios boot and 1:3 for uefi boot
isdeletable (osimage.isdeletable)
A flag to indicate whether this image profile can be deleted. This attribute is only used by PCM.
kerneldir (linuximage.kerneldir)
The directory name where the 3rd-party kernel is stored. It is used for diskless image only.
kernelver (linuximage.kernelver)
The version of linux kernel used in the linux image. If the kernel version is not set, the default kernel in rootimgdir will be used
kitcomponents (osimage.kitcomponents)
List of Kit Component IDs assigned to this OS Image definition.
krpmver (linuximage.krpmver)
The rpm version of kernel packages (for SLES only). If it is not set, the default rpm version of kernel packages will be used.
lpp_source (nimimage.lpp_source)
The name of the NIM lpp_source resource.
mksysb (nimimage.mksysb)
The name of a NIM mksysb resource.
netdrivers (linuximage.netdrivers)
The ethernet device drivers of the nodes which will use this linux image, at least the device driver for the nodes’ installnic should be included
nimmethod (nimimage.nimmethod)
The NIM install method to use, (ex. rte, mksysb).
nimtype (nimimage.nimtype)
The NIM client type- standalone, diskless, or dataless.
nodebootif (linuximage.nodebootif)
The network interface the stateless/statelite node will boot over (e.g. eth0)
osarch (osimage.osarch)
The hardware architecture of this node. Valid values: x86_64, ppc64, x86, ia64.
osdistroname (osimage.osdistroname)
The name of the OS distro definition. This attribute can be used to specify which OS distro to use, instead of using the osname,osvers,and osarch attributes. For *kit commands, the attribute will be used to read the osdistro table for the osname, osvers, and osarch attributes. If defined, the osname, osvers, and osarch attributes defined in the osimage table will be ignored.
osname (osimage.osname)
Operating system name- AIX or Linux.
osupdatename (osimage.osupdatename)
A comma-separated list of OS distro updates to apply to this osimage.
osvers (osimage.osvers)
The Linux operating system deployed on this node. Valid values: rhels*,rhelc*, rhas*,centos*,SL*, fedora*, sles* (where * is the version #).
otherifce (linuximage.otherifce)
Other network interfaces (e.g. eth1) in the image that should be configured via DHCP
otherpkgdir (linuximage.otherpkgdir)
The base directory where the non-distro packages are stored. Only 1 local directory supported at present.
otherpkglist (linuximage.otherpkglist)
The fully qualified name of the file that stores non-distro package lists that will be included in the image. It could be set multiple paths.The multiple paths must be seperated by ”,”.
otherpkgs (nimimage.otherpkgs)
One or more comma separated installp or rpm packages. The rpm packages must have a prefix of ‘R:’, (ex. R:foo.rpm)
paging (nimimage.paging)
The name of the NIM paging resource.
partitionfile (linuximage.partitionfile, winimage.partitionfile)
The path of the configuration file which will be used to partition the disk for the node. For stateful osimages,two types of files are supported: “<partition file absolute path>” which contains a partitioning definition that will be inserted directly into the generated autoinst configuration file and must be formatted for the corresponding OS installer (e.g. kickstart for RedHat, autoyast for SLES, pressed for Ubuntu). “s:<partitioning script absolute path>” which specifies a shell script that will be run from the OS installer configuration file %pre section; the script must write the correct partitioning definition into the file /tmp/partitionfile on the node which will be included into the configuration file during the install process. For statelite osimages, partitionfile should specify “<partition file absolute path>”; see the xCAT Statelite documentation for the xCAT defined format of this configuration file.For Ubuntu, besides “<partition file absolute path>” or “s:<partitioning script absolute path>”, the disk name(s) to partition must be specified in traditional, non-devfs format, delimited with space, it can be specified in 2 forms: “d:<the absolute path of the disk name file>” which contains the disk name(s) to partition and “s:d:<the absolute path of the disk script>” which runs in pressed/early_command and writes the disk names into the “/tmp/boot_disk” . To support other specific partition methods such as RAID or LVM in Ubuntu, some additional preseed values should be specified, these values can be specified with “c:<the absolute path of the additional pressed config file>” which contains the additional pressed entries in “d-i ...” form and “s:c:<the absolute path of the additional pressed config script>” which runs in pressed/early_command and set the preseed values with “debconf-set”. The multiple values should be delimited with comma ”,”
or
The path of partition configuration file. Since the partition configuration for bios boot mode and uefi boot mode are different, this configuration file can include both configurations if you need to support both bios and uefi mode. Either way, you must specify the boot mode in the configuration. Example of partition configuration file: [BIOS]xxxxxxx[UEFI]yyyyyyy. To simplify the setting, you also can set installto in partitionfile with section like [INSTALLTO]0:1
permission (linuximage.permission)
The mount permission of /.statelite directory is used, its default value is 755
pkgdir (linuximage.pkgdir)
The name of the directory where the distro packages are stored. It could be set multiple paths.The multiple paths must be seperated by ”,”. The first path in the value of osimage.pkgdir must be the OS base pkg dir path, such as pkgdir=/install/rhels6.2/x86_64,/install/updates . In the os base pkg path, there are default repository data. And in the other pkg path(s), the users should make sure there are repository data. If not, use “createrepo” command to create them. For ubuntu, multiple mirrors can be specified in the pkgdir attribute, the mirrors must be prefixed by the protocol(http/ssh) and delimited with ”,” between each other.
pkglist (linuximage.pkglist)
The fully qualified name of the file that stores the distro packages list that will be included in the image. Make sure that if the pkgs in the pkglist have dependency pkgs, the dependency pkgs should be found in one of the pkgdir
postbootscripts (osimage.postbootscripts)
Comma separated list of scripts that should be run on this after diskfull installation or diskless boot. On AIX these scripts are run during the processing of /etc/inittab. On Linux they are run at the init.d time. xCAT automatically adds the scripts in the xcatdefaults.postbootscripts attribute to run first in the list. See the site table runbootscripts attribute.
postinstall (linuximage.postinstall)
The fully qualified name of the script file that will be run at the end of the genimage command. It could be set multiple paths.The multiple paths must be seperated by ”,”. It is used for diskless image only.
postscripts (osimage.postscripts)
Comma separated list of scripts that should be run on this image after diskfull installation or diskless boot. For installation of RedHat, CentOS, Fedora, the scripts will be run before the reboot. For installation of SLES, the scripts will be run after the reboot but before the init.d process. For diskless deployment, the scripts will be run at the init.d time, and xCAT will automatically add the list of scripts from the postbootscripts attribute to run after postscripts list. For installation of AIX, the scripts will run after the reboot and acts the same as the postbootscripts attribute. For AIX, use the postbootscripts attribute. See the site table runbootscripts attribute. Support will be added in the future for the postscripts attribute to run the scripts before the reboot in AIX.
profile (osimage.profile)
The node usage category. For example compute, service.
provmethod (osimage.provmethod)
The provisioning method for node deployment. The valid values are install, netboot,statelite,boottarget,dualboot,sysclone. If boottarget is set, you must set linuximage.boottarget to the name of the boottarget definition. It is not used by AIX.
resolv_conf (nimimage.resolv_conf)
The name of the NIM resolv_conf resource.
root (nimimage.root)
The name of the NIM root resource.
rootfstype (osimage.rootfstype)
The filesystem type for the rootfs is used when the provmethod is statelite. The valid values are nfs or ramdisk. The default value is nfs
rootimgdir (linuximage.rootimgdir)
The directory name where the image is stored. It is generally used for diskless image. it also can be used in sysclone environment to specify where the image captured from golden client is stored. in sysclone environment, rootimgdir is generally assigned to some default value by xcat, but you can specify your own store directory. just one thing need to be noticed, wherever you save the image, the name of last level directory must be the name of image. for example, if your image name is testimage and you want to save this image under home directoy, rootimgdir should be assigned to value /home/testimage/
script (nimimage.script)
The name of a NIM script resource.
serverrole (osimage.serverrole)
The role of the server created by this osimage. Default roles: mgtnode, servicenode, compute, login, storage, utility.
shared_home (nimimage.shared_home)
The name of the NIM shared_home resource.
shared_root (nimimage.shared_root)
A shared_root resource represents a directory that can be used as a / (root) directory by one or more diskless clients.
spot (nimimage.spot)
The name of the NIM SPOT resource.
synclists (osimage.synclists)
The fully qualified name of a file containing a list of files to synchronize on the nodes. Can be a comma separated list of multiple synclist files. The synclist generated by PCM named /install/osimages/<imagename>/synclist.cfm is reserved for use only by PCM and should not be edited by the admin.
template (linuximage.template, winimage.template)
The fully qualified name of the template file that will be used to create the OS installer configuration file for stateful installations (e.g. kickstart for RedHat, autoyast for SLES).
or
The fully qualified name of the template file that is used to create the windows unattend.xml file for diskful installation.
tmp (nimimage.tmp)
The name of the NIM tmp resource.
usercomment (linuximage.comments, nimimage.comments)
Any user-written notes.
or
Any user-provided notes.
winpepath (winimage.winpepath)
The path of winpe which will be used to boot this image. If the real path is /tftpboot/winboot/winpe1/, the value for winpepath should be set to winboot/winpe1
mkdef(1), chdef(1), lsdef(1), rmdef(1)
policy.7¶
policy - a logical object definition in the xCAT database.
policy Attributes: commands, host, name, noderange, parameters, priority, rule, time, usercomment
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
commands (policy.commands)
The list of commands that this rule applies to. Default is “*” (all commands).
host (policy.host)
The host from which users may issue the commands specified by this rule. Default is “*” (all hosts). Only all or one host is supported
name (policy.name)
The username that is allowed to perform the commands specified by this rule. Default is “*” (all users).
noderange (policy.noderange)
The Noderange that this rule applies to. Default is “*” (all nodes). Not supported with the *def commands.
parameters (policy.parameters)
A regular expression that matches the command parameters (everything except the noderange) that this rule applies to. Default is “*” (all parameters). Not supported with the *def commands.
priority (policy.priority)
The priority value for this rule. This value is used to identify this policy data object (i.e. this rule) The table is sorted on this field with the lower the number the higher the priority. For example 1.0 is higher priority than 4.1 is higher than 4.9.
rule (policy.rule)
Specifies how this rule should be applied. Valid values are: allow, accept, trusted. Allow or accept will allow the user to run the commands. Any other value will deny the user access to the commands. Trusted means that once this client has been authenticated via the certificate, all other information that is sent (e.g. the username) is believed without question. This authorization should only be given to the xcatd on the management node at this time.
time (policy.time)
Time ranges that this command may be executed in. This is not supported.
usercomment (policy.comments)
Any user-written notes.
mkdef(1), chdef(1), lsdef(1), rmdef(1)
rack.7¶
rack - a logical object definition in the xCAT database.
rack Attributes: displayname, height, num, rackname, room, usercomment
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
displayname (rack.displayname)
Alternative name for rack. Only used by PCM.
height (rack.height)
Number of units which can be stored in the rack.
num (rack.num)
The rack number.
rackname (rack.rackname)
The rack name.
room (rack.room)
The room in which the rack is located.
usercomment (rack.comments)
Any user-written notes.
mkdef(1), chdef(1), lsdef(1), rmdef(1)
route.7¶
route - a logical object definition in the xCAT database.
route Attributes: gateway, ifname, mask, net, routename, usercomment
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
gateway (routes.gateway)
The gateway that routes the ip traffic from the mn to the nodes. It is usually a service node.
ifname (routes.ifname)
The interface name that facing the gateway. It is optional for IPv4 routes, but it is required for IPv6 routes.
mask (routes.mask)
The network mask.
net (routes.net)
The network address.
routename (routes.routename)
Name used to identify this route.
usercomment (routes.comments)
Any user-written notes.
mkdef(1), chdef(1), lsdef(1), rmdef(1)
site.7¶
site - a logical object definition in the xCAT database.
site Attributes: installdir, master, xcatdport
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
installdir (site.value)
The installation directory
master (site.value)
The management node
xcatdport (site.value)
Port used by xcatd daemon on master
mkdef(1), chdef(1), lsdef(1), rmdef(1)
zone.7¶
zone - a logical object definition in the xCAT database.
zone Attributes: defaultzone, sshbetweennodes, sshkeydir, usercomment, zonename
Logical objects of this type are stored in the xCAT database in one or more tables. Use the following commands to manipulate the objects: mkdef, chdef, lsdef, and rmdef. These commands will take care of knowing which tables the object attributes should be stored in. The attribute list below shows, in parentheses, what tables each attribute is stored in.
defaultzone (zone.defaultzone)
If nodes are not assigned to any other zone, they will default to this zone. If value is set to yes or 1.
sshbetweennodes (zone.sshbetweennodes)
Indicates whether passwordless ssh will be setup between the nodes of this zone. Values are yes/1 or no/0. Default is yes.
sshkeydir (zone.sshkeydir)
Directory containing the shared root ssh RSA keys.
usercomment (zone.comments)
Any user-provided notes.
zonename (zone.zonename)
The name of the zone.
mkdef(1), chdef(1), lsdef(1), rmdef(1)
man8¶
chtab.8¶
chtab - Add, delete or update rows in the database tables.
chtab [-h| –help]
chtab [-v| –version]
*chtab [keycolname=keyvalue] [tablename.colname=newvalue] *
*chtab [keycolname=keyvalue] [tablename.colname+=newvalue] *
*chtab -d [keycolname=keyvalue] [tablename.colname=newvalue] *
The chtab command adds, deletes or updates the attribute value in the specified table.column for the specified keyvalue. Normally, the given value will completely replace the current attribute value. But if “+=” is used instead of “=”, the specified value will be appended to the coma separated list of the attribute, if it is not already there.
- To add a node=node1 to the nodelist table with groups=all:
*chtab node=node1 nodelist.groups=all *
- To add a keyword (tftpdir) and value (/tftpboot) to the site table:
*chtab key=tftpdir site.value=/tftpboot *
- To add node1 to the nodetype table with os=rhel5:
chtab node=node1 nodetype.os=rhel5
- To change node1 in nodetype table setting os=sles:
chtab node=node1 nodetype.os=sles
- To change node1 by appending otherpkgs to the postbootscripts field in the postscripts table:
chtab node=node1 postscripts.postbootscripts+=otherpkgs
- To delete node1 from nodetype table:
chtab -d node=node1 nodetype
/opt/xcat/bin/chtab
tabdump(8)|tabdump.8, tabedit(8)|tabedit.8
copycds-cdrom.8¶
copycds-cdrom [copycds options] [drive]
copycds-cdrom is a wrapper scripts for copycds to copy from physical CD/DVD-ROM drives located on the management server.
[copycds options] are passed unmolested to copycds.
If [drive] is not specified, /dev/cdrom is assumed.
The copycds command copies all contents of Distribution CDs or Service Pack CDs to the install directory as designated in the site table attribute: installdir.
copycds(8)|copycds.8
Isaac Freeman <ifreeman@us.ibm.com>
copycds.8¶
copycds - Copies Linux distributions and service levels from DVDs/ISOs to the xCAT /install directory.
copycds [{-n|--name|--osver}=distroname] [{-a|--arch}=architecture] [{-p|--path}=ospkgpath] [-o | --noosimage] [-w | --nonoverwrite] {iso|device-path} ...
copycds [-i | --inspection] {iso|device-path}
copycds [-h | --help]
The copycds command copies all contents of Distribution DVDs/ISOs or Service Pack DVDs/ISOs to a destination directory. The destination directory can be specified by the -p option. If no path is specified, the default destination directory will be formed from the installdir site attribute and the distro name and architecture, for example: /install/rhels6.3/x86_64. The copycds command can copy from one or more ISO files, or the CD/DVD device path.
You can specify -i or –inspection option to check whether the DVDs/ISOs can be recognized by xCAT. If recognized, the distribution name, architecture and the disc no (the disc sequence number of DVDs/ISOs in multi-disk distribution) of the DVD/ISO is displayed. If xCAT doesn’t recognize the DVD/ISO, you must manually specify the distro name and architecture using the -n and -a options. This is sometimes the case for distros that have very recently been released, and the xCAT code hasn’t been updated for it yet.
You can get xCAT to recognize new DVDs/ISOs by adding them to /opt/xcat/lib/perl/xCAT/data/discinfo.pm and reloading xcatd (service xcatd reload).
{-n|--name|--osver}=distroname
The linux distro name and version that the ISO/DVD contains. Examples: rhels6.3, sles11.2, fedora9. Note the ‘s’ in rhels6.3 which denotes the Server version of RHEL, which is typically used.
{-a|--arch}=architecture
The architecture of the linux distro on the ISO/DVD. Examples: x86, x86_64, ppc64, s390x.
{-p|--path}=ospkgpath
The destination directory to which the contents of ISO/DVD will be copied. When this option is not specified, the default destination directory will be formed from the installdir site attribute and the distro name and architecture, for example: /install/rhel6.3/x86_64. This option is only supported distributions of sles,redhat and windows.
{-i|--inspection}
Check whether xCAT can recognize the DVDs/ISOs in the argument list, but do not copy the disc. Displays the os distribution name, architecture and disc no of each recognized DVD/ISO. This option only supported for distributions of sles,redhat and windows.
{-o|--noosimage}
Do not create the default osimages based on the osdistro copied in. By default, copycds will create a set of osimages based on the osdistro.
{-w|--nonoverwrite}
Complain and exit if the os disc has already been copied in. By default, copycds will overwrite the os disc already copied in.
0: The command completed successfully. For the –inspection option, the ISO/DVD have been recognized successfully
Nonzero: An Error has occurred. For the –inspection option, the ISO/DVD cannot be recognized
*
To copy the RPMs from a set of ISOs that represent the DVDs of a distro:
copycds dvd1.iso dvd2.iso
*
To copy the RPMs from a physical DVD to /depot/kits/3 directory:
copycds -p /depot/kits/3 /dev/dvd
*
To copy the RPMs from a DVD ISO of a very recently released distro:
copycds -n rhels6.4 -a x86_64 dvd.iso
*
To check whether a DVD ISO can be recognized by xCAT and display the recognized disc info:
copycds -i /media/RHEL/6.2/RHEL6.2-20111117.0-Server-ppc64-DVD1.isoOutput will be similar to:
OS Image:/media/RHEL/6.2/RHEL6.2-20111117.0-Server-ppc64-DVD1.iso DISTNAME:rhels6.2 ARCH:ppc64 DISCNO:1For the attributes that are not recognized, the value will be blank.
*
To copy the packages from a supplemental DVD ISO file:
copycds -n /isodir/RHEL6.5/RHEL6.5-Supplementary-20131114.2-Server-ppc64-DVD1.iso -n rhels6.5-suppAlso, remember to add the new directory to your osimage definition:
chdef -t osimage myosimage -p pkgdir=/install/rhels6.5-supp/ppc64
nodeset(8)|nodeset.8, site(5)|site.5, nodetype(5)|nodetype.5
makeconservercf.8¶
makeconservercf - creates the conserver configuration file from info in the xCAT database
makeconservercf [-V|--verbose] [-d|--delete] noderange
makeconservercf [-V|--verbose] [-l|--local] [noderange]
makeconservercf [-V|--verbose] [-c|--conserver] [noderange]
makeconservercf [-V|--verbose] noderange [-t|--trust] hosts
makeconservercf [-h|--help|-v|--version]
The makeconservercf command will write out the /etc/conserver.cf, using information from the nodehm table and related tables (e.g. mp, ipmi, ppc). Normally, makeconservercf will write all nodes to the conserver.cf file. If a noderange is specified, it will write on those nodes to the file. In either case, if a node does not have nodehm.cons set, it will not be written to the file.
In the case of a hierarchical cluster (i.e. one with service nodes) makeconservercf will determine which nodes will have their consoles accessed from the management node and which from a service node (based on the nodehm.conserver attribute). The /etc/conserver.cf file will be created accordingly on all relevant management/service nodes. If -l is specified, it will only create the local file.
-d|--delete
Delete rather than add or refresh the nodes specified as a noderange.
-c|--conserver
Only set up the conserver on the conserver host. If no conserver host is set for nodes, the conserver gets set up only on the management node.
-l|--local
Only run makeconservercf locally and create the local /etc/conserver.cf. The default is to also run it on all service nodes, if there are any.
-t|--trust hosts
Add additional trusted hosts into /etc/conserver.cf. The hosts are comma separated list of ip addresses or host names.
-v|--version
Display version.
-V|--verbose
Verbose mode.
-h|--help
Display usage message.
To create conserver configuration for all the nodes.
makeconservercfTo create conserver configuration for nodes node01-node10.
makeconservercf node01-node10
To remove conserver configuration for node01.
makeconservercf -d node01
rcons(1)|rcons.1
makedhcp.8¶
makedhcp - Creates and updates DHCP configuration files.
makedhcp -n [-l | --localonly]
makedhcp -a [-l | --localonly]
makedhcp -a -d [-l | --localonly]
makedhcp -d noderange [-l | --localonly]
makedhcp noderange [-s statements] [-l | --localonly]
makedhcp -q noderange
makedhcp [-h|-***-help*]
The makedhcp command creates and updates the DHCP configuration on the management node and service nodes. The makedhcp command is supported for both Linux and AIX clusters.
- Start by filling out the networks(5)|networks.5 table properly.
- Then use the makedhcp -n option to create a new dhcp configuration file. You can set the site table, dhcplease attribute to the lease time for the dhcp client. The default value is 43200.
- Next, get the node IP addresses and MACs defined in the xCAT database. Also, get the hostnames and IP addresses pushed to /etc/hosts (using makehosts(8)|makehosts.8) and to DNS (using makedns(8)|makedns.8).
- Then run makedhcp with a noderange or the -a option. This will inject into dhcpd configuration data pertinent to the specified nodes. On linux, the configuration information immediately takes effect without a restart of DHCP.
If you need to delete node entries from the DHCP configuration, use the -d flag.
-n
Create a new dhcp configuration file with a network statement for each network the dhcp daemon should listen on. (Which networks dhcpd should listen on can be controlled by the dhcpinterfaces attribute in the site(5)|site.5 table.) The makedhcp command will automatically restart the dhcp daemon after this operation. This option will replace any existing configuration file (making a backup of it first). For Linux systems the file will include network entries as well as certain general parameters such as a dynamic range and omapi configuration. For AIX systems the file will include network entries. On AIX systems, if there are any non-xCAT entries in the existing configuration file they will be preserved and added to the end of the new configuration file.
-a
Define all nodes to the DHCP server. (Will only add nodes that can be reached, network-wise, by this DHCP server.) The dhcp daemon does not have to be restarted after this. On AIX systems makedhcp will not add entries for cluster nodes that will be installed using NIM. The entries for these nodes will be managed by NIM.
noderange
Add the specified nodes to the DHCP server configuration.
-s statements
For the input noderange, the argument will be interpreted like dhcp configuration file text.
-d noderange
Delete node entries from the DHCP server configuration. On AIX, any entries created by NIM will not be removed.
-a -d
Delete all node entries, that were added by xCAT, from the DHCP server configuration.
-l | --localonly
Configure dhcpd on the local machine only. Without this option, makedhcp will also send this operation to any service nodes that service the nodes in the noderange.
-q noderange
Query the node entries from the DHCP server configuration. On AIX, any entries created by NIM will not be listed.
-h|--help
Display usage message.
Create a new DHCP configuration file and add the network definitions:
makedhcp -n
Define all nodes to the dhcp server:
makedhcp -a
Note: This does not add nodes that will be installed with AIX/NIM.
Will cause dhcp on the next request to set root-path appropriately for only node5. Note some characters (e.g. ”) must be doubly escaped (once for the shell, and once for the OMAPI layer).
makedhcp node5 -s 'option root-path \"172.16.0.1:/install/freebsd6.2/x86_64\";'
Query a node from the DHCP server.
# makedhcp -q node01 node01: ip-address = 91.214.34.156, hardware-address = 00:00:c9:c6:6c:42
DHCP configuration files:
[AIX] /etc/dhcpsd.cnf
[SLES] /etc/dhcpd.conf
[RH] /etc/dhcp/dhcpd.conf
noderange(3)|noderange.3
makedns.8¶
makedns - sets up domain name services (DNS).
makedns [-V|-**-verbose] [-e|--external] [*-n*|*-**-new*] [noderange]
makedns [-V|-**-verbose] [-e|--external] [*-d*|*-**-delete* noderange]
makedns configures a DNS server on the system you run it on, which is typically the xCAT management node.
The list of nodes to include comes from either the noderange provided on the command line or the entries in the local /etc/hosts files.
There are several bits of information that must be included in the xCAT database before running this command.
You must set the forwarders attributes in the xCAT site definition.
The forwarders value should be set to the IP address of one or more nameservers at your site that can resolve names outside of your cluster. With this set up, all nodes ask the local nameserver to resolve names, and if it is a name that the MN DNS does not know about, it will try the forwarder names.
An xCAT network definition must be defined for each network used in the cluster. The net and mask attributes will be used by the makedns command.
A network domain and nameservers values must be provided either in the network definiton corresponding to the node or in the site definition.
Only entries in /etc/hosts or the hosts specified by noderange that have a corresponding xCAT network definition will be added to DNS.
By default, makedns sets up the named service and updates the DNS records on the local system (management node). If the -e flag is specified, it will also update the DNS records on any external DNS server that is listed in the /etc/resolv.conf on the management node. (Assuming the external DNS server can recognize the xCAT key as authentication.)
For more information on Cluster Name Resolution: Cluster_Name_Resolution
-V | --verbose
Verbose mode.
-n | --new
Use this flag to create new named configuration and db files.
-d | --delete
Remove the DNS records.
-e | --external
Update DNS records to the external DNS server listed in /etc/resolv.conf.
Enabling the site attribute externaldns means use ‘external’ DNS by default. If setting externaldns to 1, you need NOT use -e flag in every makedns call.
noderange
A set of comma delimited node names and/or group names. See the “noderange” man page for details on additional supported formats.
1
To set up DNS for all the hosts in /etc/hosts file.
makedns
2
To set up DNS for node1.
makedns node1
3
To create a new named configuration and db files for all hosts in /etc/hosts.
makedns -n
4
To delete the DNS records for node1.
makedns -d node1
makehosts(8)|makehosts.8
makehosts.8¶
makehosts - sets up /etc/hosts from the xCAT hosts table.
makehosts [-n] [noderange] [-l | --longnamefirst] [-d] [-m | --mactolinklocal]
makehosts {-h | --help}
makehosts updates the /etc/hosts file based on information stored in the xCAT database object definitions.
The main three bits of information needed are: node hostname, node ip and network domain name.
The hostname and ip address are specified as part of the node definition.
The domain value is taken either from the xCAT network definition associated with the node or from the cluster site definition. If you are using multiple domains in the cluster you should add the domain names to the appropriate xCAT network definition.
Note: If your node hostnames and IP addresses follow a regular pattern, you can use just a few regular expressions to generate /etc/hosts using makehosts. For details on using regular expressions see the “xcatdb” man page.
If you specify additional network interfaces in your xCAT node definitions they will also be added to the /etc/hosts file. You can specify additional network interface information (NICs) using the following node attributes: nicips, nichostnamesuffixes, nictypes, niccustomscripts, nicnetworks. You can get a description of these attributes by running “lsdef -t node -h | more” or “man nics”.
-n
Completely replace the /etc/hosts file, losing any previous content. If this option is not specified, it will only replace the lines in the file that correspond to the nodes in the specified noderange.
-l | --longnamefirst
The long name of the host will appear before the short name for each host in the /etc/hosts file. The default is short name first.
-m | --mactolinklocal
Updates /etc/hosts file with IPv6 link local addresses, the link local address is generated from the mac address stored in mac table.
-d
Delete rather than create records. This will also delete any additional network interfaces (NICs) included in the node definitions.
*
Add entries to /etc/hosts for all nodes included in the xCAT node group called “compute”.
makehosts compute
*
If the xCAT hosts table contains:
"compute","|node(\d+)|1.2.3.($1+0)|","|(.*)|($1).cluster.net|",,Assuming the group “compute” contains node01, node02, etc., then in /etc/hosts they will be given IP addresses of 1.2.3.1, 1.2.3.2, etc.
hosts(5)|hosts.5, makedns(8)|makedns.8
makeknownhosts.8¶
makeknownhosts - Make a known_hosts file under $ROOTHOME/.ssh for input noderange.
makeknownhosts Replaces or removes in the known_hosts file in the $ROOTHOME/.ssh directory, the enties for the nodes from the noderange input to the command. The known_hosts file entry is built from the shared ssh host key that xCAT distributes to the installed nodes.
HMCs, AMM, switches, etc., where xCAT does not distribute the shared ssh host key, should not be put in the noderange.
To build the known_hosts entry for a node, you are only required to have the node in the database, and name resolution working for the node. You do not have to be able to access the node.
Having this file with correct entries, will avoid the ssh warning when nodes are automatically added to the known_hosts file. The file should be distributed using xdcp to all the nodes, if you want node to node communication not to display the warning.
noderange
A set of comma delimited node names and/or group names. See the “noderange” man page for details on supported formats.
-r|--remove
Only removes the entries for the nodes from the known_hosts file.
-V|--verbose
Verbose mode.
makenetworks.8¶
makenetworks - Gather cluster network information and add it to the xCAT database.
makenetworks [-h|–help ]
makenetworks [-v| –version]
makenetworks [-V|–verbose] [-d|–display]
The makenetworks command can be used to gather network information from an xCAT cluster environment and create corresponding network definitions in the xCAT database.
Every network that will be used to install a cluster node must be defined in the xCAT database.
The default behavior is to gather network information from the managment node, and any configured xCAT service nodes, and automatically save this information in the xCAT database.
You can use the “-d” option to display the network information without writing it to the database.
You can also redirect the output to a file that can be used with the xCAT mkdef command to define the networks.
For example:
makenetworks -d > mynetstanzas
cat mynetstanzas | mkdef -z
This features allows you to verify and modify the network information before writing it to the database.
When the network information is gathered a default value is created for the “netname” attribute. This is done to make it possible to use the mkdef, chdef, lsdef, and rmdef commands to manage this data.
The default naming convention is to use a hyphen separated “net” and “mask” value with the ”.” replace by “_”. (ex. “8_124_47_64-255_255_255_0”)
You can also modify the xCAT “networks” database table directly using the xCAT tabedit command.
tabedit networks
Note: The makenetworks command is run automatically when xCAT is installed on a Linux management node.
-d|--display Display the network definitions but do not write to the definitions to the xCAT database. The output will be in stanza file format and can be redirected to a stanza file that can be used with mkdef or chdef commands to create or modify the network definitions.
-h | --help Display usage message.
-v | --version Command Version.
-V |--verbose Verbose mode.
- Gather cluster network information and create xCAT network definitions.
makenetworks
- Display cluster network information but do not write the network definitions to the xCAT database.
makenetworks -d
The output would be one or more stanzas of information similar to the following. The line that ends with a colon is the value of the “netname” attribute and is the name of the network object to use with the lsdef, mkdef, chdef and rmdef commands.
- 9_114_37_0-255_255_255_0:
- objtype=network gateway=9.114.37.254 mask=255.255.255.0 net=9.114.37.0
/opt/xcat/sbin/makenetworks
makedhcp(8)|makedhcp.8
makeroutes.8¶
makeroutes - add or delete routes to/from the os route table on nodes.
makeroutes [-r | --routename r1[,r2...]]
makeroutes -d | --delete [-r | --routenames r1[,r2...]]
makeroutes noderange [-r | --routename r1[,r2...]]
makeroutes noderange -d | --delete [-r | --routenames r1[,r2...]]
makeroutes [-h --help | -v | --version]
The makeroutes command adds or deletes routes on the management node or any given nodes. The noderange specifies the nodes where the routes are to be added or removed. When the noderange is omitted, the action will be done on the management node. The -r option specifies the name of routes. The details of the routes are defined in the routes table which contians the route name, subnet, net mask and gateway. If -r option is omitted, the names of the routes found on noderes.routenames for the nodes or on site.mnroutenames for the management node will be used.
If you want the routes be automatically setup during node deployment, first put a list of route names to noderes.routenames and then add setroute script name to the postscripts.postbootscripts for the nodes.
noderange specifies the nodes where the routes are to be added or removed. If omitted, the operation will be done on the management node.
-d|--delete
Specifies to delete the given routes. If not specified, the action is to add routes.
-r|--routename
Specifies a list of comma separated route names defined in the routes table. If omitted, all routes defined in noderes.routenames for nodes or site.mnroutenames for the management node will be used.
-h|--help
Display usage message.
-v|--version
Command Version.
To add all routes from the site.mnroutenames to the os route table for the management node.
makeroutesTo add all the routes from noderes.routenames to the os route table for node1.
makeroutes node1
To add route rr1 and rr2 to the os route table for the management node.
makeroutes -r rr1,rr2
To delete route rr1 and rr2 from the os route table on node1 and node1.
makeroutes node1,node2 -d -r rr1,rr2
/opt/xcat/sbin/makeroutes
mknb.8¶
mknb - creates a network boot root image for node discovery and flashing
mknb arch
The mknb command is run by xCAT automatically, when xCAT is installed on the management node. It creates a network boot root image (used for node discovery, BMC programming, and flashing) for the same architecture that the management node is. So you normally do not need to run the mknbcommand yourself.
If you do run mknb to add custom utilities to your boot root image, and you have an xCAT Hierarchical Cluster with service nodes that each have a local /tftpboot directory (site sharedtftp=0), you will also need to copy the generated root image to each service node.
Presently, only the arch x86_64 is supported.
makedhcp(8)|makedhcp.8
nodeadd.8¶
nodeadd - Adds nodes to the xCAT cluster database.
nodeadd noderange groups=groupnames [table.column=value] [...]
nodeadd {-v | --version}
nodeadd [-? | -h | --help]
The nodeadd command adds the nodes specified in noderange to the xCAT database. It also stores the any additional attributes specified for these nodes. At least one groupname must be supplied. You should also consider specifying attributes in at least the following tables: nodehm, noderes, nodetype. See the man page for each of these for details. Also see the xcatdb man page for an overview of each table.
The nodeadd command also supports some short cut names as aliases to common attributes. See the nodels man page for details.
*
To add nodes in noderange node1-node4 with group all:
nodeadd node1-node4 groups=all
*
To add nodes in noderange node1-node4 to the nodetype table with os=rhel5:
nodeadd node1-node4 groups=all,rhel5 nodetype.os=rhel5
/opt/xcat/bin/nodeadd
nodels(1)|nodels.1, nodech(1)|nodech.1, noderange(3)|noderange.3
nodeset.8¶
nodeset - set the boot state for a noderange
nodeset [noderange] [boot|\ *install*\ |\ *stat*\ |\ *iscsiboot*\ |\ *netboot*\ |\ *statelite*\ |\ *offline*\ |\ *runcmd=bmcsetup*\ |\ *osimage[=<imagename*\ >]|shell|shutdown]
nodeset noderange osimage=<imagename> [-*-noupdateinitrd*] [*-**-ignorekernelchk*]
nodeset sets the next boot state for a single or range of nodes or groups. It tells xCAT what you want to happen the next time the nodes are booted up. See noderange(3)|noderange.3. nodeset accomplishes this by changing the network boot files. Each xCAT node always boots from the network and downloads a boot file with instructions on what action to take next.
nodeset will manipulate the boot configuration files of yaboot and pxelinux.0.
Assume that /tftpboot is the root for tftpd (set in site(5)|site.5).
nodeset for pxe makes changes to /tftpboot/pxelinux.cfg/{node hex ip}
nodeset for yaboot makes changes to /tftpboot/etc/{node hex ip}
nodeset only sets the next boot state, but does not reboot.
nodeset is called by rinstall and winstall and is also called by the installation process remotely to set the boot state back to “boot”.
A user can supply their own scripts to be run on the mn or on the service node (if a hierarchical cluster) for a node when the nodeset command is run. Such scripts are called prescripts. They should be copied to /install/prescripts dirctory. A table called prescripts is used to specify the scripts and their associated actions. The scripts to be run at the beginning of the nodeset command are stored in the ‘begin’ column of prescripts table. The scripts to be run at the end of the nodeset command are stored in the ‘end’ column of prescripts table. You can run ‘tabdump prescripts -d’ command for details. The following two environment variables will be passed to each script: NODES contains all the names of the nodes that need to run the script for and ACTION contains the current nodeset action. If #xCAT setting:MAX_INSTANCE=number is specified in the script, the script will get invoked for each node in parallel, but no more than number of instances will be invoked at at a time. If it is not specified, the script will be invoked once for all the nodes.
boot
Instruct network boot loader to be skipped, generally meaning boot to hard disk
install | install=profilename | install=<os-architecture-profilename>
Prepare server for installing a node according to the table configuration. This will typically mean an unattended install file is created (i.e. kickstart or autoyast), install resources will be verified in tftp and the bulk transfer method (http, ftp, or nfs for example), and the boot loader configuration file will be written. Note: this option has been deprecated and will be removed in a future release. You should change to using the osimage option instead.
netboot | netboot=profilename | netboot=<os-architecture-profilename>
Prepare server for deploying diskless nodes. With the diskless solution, the os image will reside in the RAM of the node. Note: this option has been deprecated and will be removed in a future release. You should change to using the osimage option instead.
statelite | statelite=profilename | statelite=<os-architecture-profilename>
Prepare server for deploying statelite nodes. Statelite provides an efficient and flexible diskless solution because most of the OS image is NFS mounted read-only, but a configurable list of directories and files can be read-write. The read-write files can either be persistent across reboots, or volatile (restoring to pristine state after reboot). The configuration can be done through litefile, litetree and statelite tables. Note: this option has been deprecated and will be removed in a future release. You should change to using the osimage option instead.
offline
Cleanup the current pxe/tftp boot configuration files for the nodes requested
osimage | osimage=<imagename>
Prepare server for installing a node using the specified os image. The os image is defined in the osimage table and linuximage table. If the <imagename> is omitted, the os image name will be obtained from nodetype.provmethod for the node.
--noupdateinitrd
Skip the rebuilding of initrd when the ‘netdrivers’, ‘drvierupdatesrc’ or ‘osupdatename’ were set for injecting new drviers to initrd. But, the geninitrd command should be run to rebuild the initrd for new drivers injecting. This is used to improve the performance of nodeset command.
--ignorekernelchk
Skip the kernel version checking when injecting drivers from osimage.driverupdatesrc. That means all drivers from osimage.driverupdatesrc will be injected to initrd for the specific target kernel.
runimage=<task>>
If you would like to run a task after deployment, you can define that task with this attribute.
stat
Display the current boot loader config file description for the nodes requested
runcmd=bmcsetup
This instructs the node to boot to the xCAT nbfs environment and proceed to configure BMC for basic remote access. This causes the IP, netmask, gateway, username, and password to be programmed according to the configuration table.
shell
This instructs tho node to boot to the xCAT genesis environment, and present a shell prompt on console. The node will also be able to be sshed into and have utilities such as wget, tftp, scp, nfs, and cifs. It will have storage drivers available for many common systems.
shutdown
To make the node to get into power off status. This status only can be used after runcmd and runimage to power off the node after the performing of operations.
-h | --help
Print help.
-v | --version
Print version.
noderes table - xCAT node resources file. See noderes(5)|noderes.5 for further details.
nodetype table - xCAT node installation type file. See nodetype(5)|nodetype.5 for fur- ther details. This is used to determine the node installation image type.
site table - xCAT main configuration file. See site(5)|site.5 for further details. This is used to determine the location of the TFTP root directory and the TFTP xCAT subdirectory. /tftpboot and /tftpboot/xcat is the default.
*
To setup to install mycomputeimage on the compute node group.
nodeset compute osimage=mycomputeimage
*
To run http://$master/image.tgz after deployment:
nodeset $node runimage=http://$MASTER/image.tgznodeset
noderange(3)|noderange.3, nodels(1)|nodels.1, nodestat(1)|nodestat.1, rinstall(8)|rinstall.8, makedhcp(8)|makedhcp.8, osimage(7)|osimage.7
rescanplugins.8¶
rescanplugins - Notifies xcatd to rescan the plugin directory
rescanplugins
rescanplugins {-h | --help}
rescanplugins {-v | --version}
rescanplugins [-s | --servicenodes]
rescanplugins notifies the xcatd daemon to rescan the plugin directory and update its internal command handlers hash. This command should be used when plugins have been added or removed from the xCAT plugin directory (/opt/xcat/lib/perl/xCAT_plugin) or if the contents of the handled_commands subroutine in an existing plugin has changed.
If rescanplugins is called as a subrequest from another command, the xcatd command handlers hash changes will not be available to that command’s process. Only subsequent command calls will see the updates.
-h|--help
Displays the usage message.
-v|--version
Displays the release version of the code.
-s|--servicenodes
Process the rescanplugins on the management node and on all service nodes. The rescanplugins command will be sent to the xcatd daemon on all nodes defined in the servicenode table. The default is to only run on the management node.
*
To rescan the plugins only on the xCAT Management Node:
rescanplugins
*
To rescan the plugins on the xCAT Management Node and on all service nodes:
rescanplugins -s
rinstall.8¶
rinstall - Begin OS provision on a noderange
rinstall [-o|-*-osver*] [*-p*|*--profile*] [*-a*|*--arch*] [*-c*|*-**-console*] [noderange]
rinstall is a convenience command that will change tables as requested for operating system version, profile, and architecture, call nodeset to modify the network boot configuration, call rsetboot net to set the next boot over network (only support nodes with “nodetype.mgt=ipmi”, for other nodes, make sure the correct boot order has been set before rinstall), and rpower to begin a boot cycle.
If [-O|–osimage] is specified or nodetype.provmethod=osimage is set, provision the noderange with the osimage specified/configured, ignore the table change options if specified.
If -c is specified, it will then run rcons on the node. This is allowed only if one node in the noderange. If need consoles on multiple nodes , see winstall(8)|winstall.8.
-h | --help
Display usage message.
-v | --version
Display version.
-o | --osver
Specifies which os version to provision. If unspecified, the current node os setting is used. Will be ignored if [-O|–osimage] is specified or nodetype.provmethod=osimage.
-p | --profile
Specifies what profile should be used of the operating system. If not specified the current node profile setting is used. Will be ignored if [-O|–osimage] is specified or nodetype.provmethod=osimage.
-a | --arch
Specifies what architecture of the OS to provision. Typically this is unneeded, but if provisioning between x86_64 and x86 frequently, this may be a useful flag. Will be ignored if [-O|–osimage] is specified or nodetype.provmethod=osimage.
-O | --osimage
Specifies the osimage to provision.
-c | --console
Requests that rinstall runs rcons once the provision starts. This will only work if there is only one node in the noderange. See winstall(8)|winstall.8 for starting nsoles on multiple nodes.
rinstall node1-node20
Provison nodes 1 through 20, using their current configuration.
rinstall node1-node20 -o rhels5.1 -p compute
Provision nodes 1 through 20, forcing rhels5.1 and compute profile.
rinstall node1-node20 -O rhels6.4-ppc64-netboot-compute
Provision nodes 1 through 20 with the osimage rhels6.4-ppc64-netboot-compute.
rinstall node1 -c
Provisoon node1 and start a console to monitor the process.
noderange(3)|noderange.3, winstall(8)|winstall.8, rcons(1)|rcons.1
rmosdistro.8¶
rmosdistro [-a | --all] [-f|--force] osdistroname [osdistroname2 ...]
rmosdistro [-h | --help]
The rmosdistro command removes the specified OS Distro that was created by copycds. To delete all OS Distro entries, please specify [-a|--all]. If the specified OS Distro is referenced by some osimage, [-f|force] can be used to remove it.
The OS Distro names to delete, delimited by blank space.
{-a|--all}
If specified, try to delete all the OS Distros.
{-f|--force}
Remove referenced OS Distros, never prompt.
{-h|--help}
Show info of rmosdistro usage.
- Zero:
- The command completed successfully.
- Nonzero:
- An Error has occurred.
*
To remove OS Distro “rhels6.2-ppc64” and “sles11.2-ppc64”:
rmosdistro rhels6.2-ppc64 sles11.2-ppc64
*
To remove OS Distro “rhels6.2-ppc64”, regardless of whether is referenced by any osimage:
rmosdistro -f rhels6.2-ppc64
*
To remove all OS Distros:
rmosdistro -a
runsqlcmd.8¶
runsqlcmd -Runs sql command files against the current xCAT database.
runsqlcmd
runsqlcmd {-h | --help}
runsqlcmd {-v | --version}
runsqlcmd {-d | --dir directory_path} {-V | --verbose}
runsqlcmd {-f | --files list of files} {-V | --verbose}
runsqlcmd {-V | --verbose} {sql statement}
The runsqlcmd routine, runs the sql statements contained in the *.sql files as input to the command against the current running xCAT database. Only DB2,MySQL and PostgreSQL databases are supported. SQLite is not supported. If no directory or filelist is provided, the default /opt/xcat/lib/perl/xCAT_schema directory is used. If the directory is input with the -d flag, that directory will be used. If a comma separated list of files is input with the -f flag, those files will be used.
-h|--help
Displays the usage message.
-v|--version
Displays current code version.
-V|--verbose
Displays extra debug information.
-d|--dir
To use a directory other than the default directory, enter the directory path here.
-f|--files
Comma separated list of files (full path), wildcard (*) can be used.
File format
The files must be of the form <name>.sql or <name>_<database>.sql where
<database> is mysql,pgsql, or db2. Files must have permission 0755.
sql statement
Quoted sql statement syntax appropriate for the current database.
*
To run the database appropriate *.sql files in /opt/xcat/lib/perl/xCAT_schema :
runsqlcmd
*
To run the database appropriate *.sql files in /tmp/mysql:
runsqlcmd -d /tmp/mysql
*
To run the database appropriate *.sql files in the input list:
runsqlcmd -f “/tmp/mysql/test\,/tmp/mysql/test1\*”*
*
To checkout one DB2 sql file:
runsqlcmd -f /tmp/db2/test_db2.sql
*
To run the following command to the database:
runsqlcmd “Select \ from site;”*
setupiscsidev.8¶
setupiscsidev - creates a LUN for a node to boot up with, using iSCSI
The setupiscsidev command will create a LUN on the management node (or service node) for each node specified. The LUN device can then be used by the node as an iSCSI device so the node can boot diskless, stateful.
-s|--size
The size of the LUN that should be created. Default is 4096.
-v|--version
Display version.
-h|--help
Display usage message.
nodeset(8)|nodeset.8
tabch.8¶
tabch - Add, delete or update rows in the database tables.
tabch [-h| –help]
tabch [-v| –version]
*tabch [keycolname=keyvalue] [tablename.colname=newvalue] *
*tabch [keycolname=keyvalue] [tablename.colname+=newvalue] *
*tabch -d [keycolname=keyvalue] [tablename.colname=newvalue] *
The tabch command adds, deletes or updates the attribute value in the specified table.column for the specified keyvalue. The difference between tabch and chtab is tabch runs as a plugin under the xcatd daemon. This give the additional security of being authorized by the daemon. Normally, the given value will completely replace the current attribute value. But if “+=” is used instead of “=”, the specified value will be appended to the coma separated list of the attribute, if it is not already there.
- To add a node=node1 to the nodelist table with groups=all:
*tabch node=node1 nodelist.groups=all *
- To add a keyword (tftpdir) and value (/tftpboot) to the site table:
*tabch key=tftpdir site.value=/tftpboot *
- To add node1 to the nodetype table with os=rhel5:
tabch node=node1 nodetype.os=rhel5
- To change node1 in nodetype table setting os=sles:
tabch node=node1 nodetype.os=sles
- To change node1 by appending otherpkgs to the postbootscripts field in the postscripts table:
tabch node=node1 postscripts.postbootscripts+=otherpkgs
- To delete node1 from nodetype table:
tabch -d node=node1 nodetype
/opt/xcat/sbin/tabch
tabdump(8)|tabdump.8, tabedit(8)|tabedit.8
tabdump.8¶
tabdump - display an xCAT database table in CSV format.
tabdump [-d] [table]
tabdump [table]
tabdump [-f filename] [table]
tabdump [-n # of records] [auditlog | eventlog]
tabdump [-w attr==val] [-w attr=~val] ...] [table]
tabdump [-w attr==val] [-w attr=~val] ...] [-f filename] [table]
tabdump [-v | -***-version*]
tabdump [-? | -h | -***-help*]
tabdump
The tabdump command displays the header and all the rows of the specified table in CSV (comma separated values) format. Only one table can be specified. If no table is specified, the list of existing tables will be displayed.
-?|-h|--help
Display usage message.
-d
Show descriptions of the tables, instead of the contents of the tables. If a table name is also specified, descriptions of the columns (attributes) of the table will be displayed. Otherwise, a summary of each table will be displayed.
-n
Shows the most recent number of entries as supplied on the -n flag from the auditlog or eventlog table.
-f
File name or path to file in which to dump the table. Without this the table is dumped to stdout. Using the -f flag allows the table to be dumped one record at a time. If tables are very large, dumping to stdout can cause problems such as running out of memory.
-w ‘attr==val’ -w ‘attr=~val’ ...
Use one or multiple -w flags to specify the selection string that can be used to select particular rows of the table. See examples.
Operator descriptions:
== Select nodes where the attribute value is exactly this value. != Select nodes where the attribute value is not this specific value. > Select nodes where the attribute value is greater than this specific value. >= Select nodes where the attribute value is greater than or equal to this specific value. < Select nodes where the attribute value is less than this specific value. <= Select nodes where the attribute value is less than or equal to this specific value. =~ Select nodes where the attribute value matches the SQL LIKE value. !~ Select nodes where the attribute value matches the SQL NOT LIKE value.
*
To display the contents of the site table:
tabdump site
*
To display the contents of the nodelist table where the groups attribute is compute :
tabdump -w ‘groups==compute’ nodelist
*
To display the contents of the nodelist table where the groups attribute is comput% where % is a wildcard and can represent any string and the status attribute is booted :
tabdump -w ‘groups=~comput%’ -w ‘status==booted’ nodelist
*
To display the records of the auditlog on date 2011-04-18 11:30:00 :
tabdump -w ‘audittime==2011-04-18 11:30:00’ auditlog
*
To display the records of the auditlog starting on 2011-04-18:
tabdump -w ‘audittime>2011-04-18 11:30:00’ auditlog
*
To display the 10 most recent entries in the auditlog:
tabdump -n 10 auditlog
*
To see what tables exist in the xCAT database:
tabdump
*
*
To display a summary description of each table:
tabdump -d
*
To display a description of each column in the nodehm table:
tabdump -d nodehm
/opt/xcat/sbin/tabdump
tabrestore(8)|tabrestore.8, tabedit(8)|tabedit.8, dumpxCATdb(1)|dumpxCATdb.1
tabedit.8¶
tabedit - view an xCAT database table in an editor and make changes.
The tabedit command opens the specified table in the user’s editor, allows them to edit any text, and then writes changes back to the database table. The table is flattened into a CSV (comma separated values) format file before giving it to the editor. After the editor is exited, the CSV file will be translated back into the database format. You may not tabedit the auditlog or eventlog because indexes will be regenerated. Use tabprune command to edit auditlog and eventlog.
TABEDITOR
The editor that should be used to edit the table, for example: vi, vim, emacs, oocalc, pico, gnumeric, nano. If TABEDITOR is not set, the value from EDITOR will be used. If EDITOR is not set, it will default to vi.
/opt/xcat/sbin/tabedit
tabrestore(8)|tabrestore.8, tabdump(8)|tabdump.8, chtab(8)|chtab.8
tabprune.8¶
tabprune - Deletes records from the eventlog,auditlog,isnm_perf,isnm_perf_sum tables.
tabprune eventlog | auditlog [-V] -i recid |-n number of records | -p percentage | -d number of days | -a
tabprune tablename -a
tabprune [-h | --help] [-v | --version]
The tabprune command is used to delete records from the auditlog,eventlog,isnm_perf,isnm_perf_sum tables. As an option, the table header and all the rows pruned from the specified table will be displayed in CSV (comma separated values) format. The all records options (-a) can be used on any xCAT table.
-h|--help
Display usage message.
-V
Verbose mode. This will cause tabprune to display the records that are being deleted from the table, in case you want to redirect them to a file to archive them.
-a
Remove all records from the input table name. This option can be used on any xCAT table.
-i recid number
Remove the records whose recid is less than the input recid number.
-n number
Remove the number of records input.
-p percent
Remove the number of records input.
-d number of days
Remove all records that occurred >= than number of days ago.
*
To remove all the records in the eventlog table:
tabprune eventlog -a
*
To remove all the records in the eventlog table saving the deleted records in eventlog.csv:
tabprune eventlog -V -a > eventlog.csv
*
To remove all the records before recid=200 in the auditlog table:
tabprune auditlog -i 200
*
To remove 400 records from the auditlog table and display the remove records:
tabprune auditlog -V -n 400
*
To remove 50% of the eventlog table:
tabprune eventlog -p 50
*
To remove all records that occurred >= 5 days ago in the eventlog:
tabprune eventlog -d 5
/opt/xcat/sbin/tabprune
tabrestore(8)|tabrestore.8, tabedit(8)|tabedit.8,tabdump(8)|tabdump.8
tabrestore.8¶
tabrestore - replaces with or adds to a xCAT database table the contents in a csv file.
tabrestore [-a] table.csv
tabrestore [-? | -h | -***-help*]
tabrestore [v | -***-version*]
The tabrestore command reads the contents of the specified file and puts its data in the corresponding table in the xCAT database. Any existing rows in that table are replaced unless the (-a) flag is used and then the rows in the file are added to the table. The file must be in csv format. It could be created by tabdump. Only one table can be specified.
This command can be used to copy the example table entries in /opt/xcat/share/xcat/templates/e1350 into the xCAT database.
-?|-h|--help
Display usage message.
-v|--version
Display version.
-a|--addrows
Add rows from the CSV file to the table instead of replacing the table with the CSV file.
*
To replace the rows in the mp table with the rows in the mp.csv file:
** tabrestore** mp.csv
The file mp.csv could contain something like:
#node,mpa,id,comments,disable "blade","|\D+(\d+)|amm(($1-1)/14+1)|","|\D+(\d+)|(($1-1)%14+1)|",,
*
To add the rows in the mp.csv file to the rows in the mp table:
** tabrestore** -a mp.csv
*
To restore database tables that we dumped with dumpxCATdb:
restorexCATdb -p <restore directory>
/opt/xcat/sbin/tabrestore
tabdump(8)|tabdump.8, tabedit(8)|tabedit.8, dumpxCATdb(1)|dumpxCATdb.1
winstall.8¶
winstall - Begin OS provision on a noderange
winstall is a convenience tool that will change attributes as requested for operating system version, profile, and architecture, call nodeset to modify the network boot configuration, call rsetboot net to set the next boot over network (only support nodes with “nodetype.mgt=ipmi”, for other nodes, make sure the correct boot order has been set before winstall), and rpower to begin a boot cycle.
If [-O|–osimage] is specified or nodetype.provmethod=osimage is set, provision the noderange with the osimage specified/configured, ignore the table change options if specified.
It will then run wcons on the nodes.
-h | --help
Display usage message.
-v | --version
Display version.
-o | --osver
Specifies which os version to provision. If unspecified, the current node os setting is used. Will be ignored if [-O|–osimage] is specified or nodetype.provmethod=osimage.
-p | --profile
Specifies what profile should be used of the operating system. If not specified the current node profile setting is used. Will be ignored if [-O|–osimage] is specified or nodetype.provmethod=osimage.
-a | --arch
Specifies what architecture of the OS to provision. Typically this is unneeded, but if provisioning between x86_64 and x86 frequently, this may be a useful flag. Will be ignored if [-O|–osimage] is specified or nodetype.provmethod=osimage.
-O | --osimage
Specifies the osimage to provision.
winstall node1-node20
Provison nodes 1 through 20, using their current configuration.
winstall node1-node20 -o rhels5.1 -p compute
Provision nodes 1 through 20, forcing rhels5.1 and compute profile.
winstall node1-node20 -O rhels6.4-ppc64-netboot-compute
Provision nodes 1 through 20 with the osimage rhels6.4-ppc64-netboot-compute.
noderange(3)|noderange.3, rinstall(8)|rinstall.8, wcons(1)|wcons.1
xcatconfig.8¶
xcatconfig - Sets up the Management Node during the xCAT install.
xcatconfig
xcatconfig {-h | --help}
xcatconfig {-v | --version}
xcatconfig {-i | --initinstall} [-V | --verbose]
xcatconfig {-u | --updateinstall} [-V | --verbose]
xcatconfig [-k | --sshkeys] [-s | --sshnodehostkeys] [-c | --credentials] [-d | --database] [-m | --mgtnode] [-t | --tunables] [-V | --verbose]
xcatconfig {-f | --force} [-V | --verbose]
xcatconfig Performs basic xCAT setup operations on an xCAT management node. This command should not be run on an xCAT Service Node, unless you are making it a Management Node. See flag description below for more details.
-h|--help
Displays the usage message.
-v|--version
Displays the release version of the code.
-V|--verbose
Displays verbose messages.
-i|--initialinstall
The install option is normally run as a post operation from the rpm xCAT.spec file during the initial install of xCAT on the Management Node. It will setup the root ssh keys, ssh node keys, xCAT credentials, initialize the datebase, export directories, start syslog and other daemons as needed after the initial install of xCAT.
-u|--updateinstall
The update install option is normally run as a post operation from the rpm xCAT.spec file during an update install of xCAT on the Management Node. It will check the setup the root ssh keys, ssh node keys, xCAT credentials, datebase, exported directories, syslog and the state of daemons needed by xCAT, after the updateinstall of xCAT. If setup is required, it will perform the operation. It will restart the necessary daemons.
-k|--sshkeys
This option will remove and regenerate the root id_rsa keys. It should only be used, if the keys are deleted or corrupted. The keys must then be distribute to the nodes by installing, running updatenode -k, or using xdsh -K option, for root to be able to ssh to the nodes without being prompted for a password. rspconfig will need to be run to distribute the key to the MM and HMCs. Any device, we need to ssh from the MN to the device will also have to be updated with the new ssh keys.
-s|--sshnodehostkeys
This option will remove and regenerate the node host ssh keys. It should only be used, if the keys are deleted or are corrupted. The keys must then be redistribute to the nodes by installing, running updatenode -k or using xdcp or pcp to copy the keys from /etc/xcat/hostkeys directory to the /etc/ssh directory on the nodes.
-c|--credentials
This option will remove all xcat credentials for root and any userids where credentials have been created. It will regenerate roots credentials, but the admin will have to add back all the userid credentials needed with the /opt/xcat/share/xcat/scripts/setup-local-client.sh <username> command. It should only be used, if they are deleted or become corrupted. The root credentials must be redistribed to the service nodes by installing the service node or using updatenode -k. makeconservercf must be rerun to pick up the new credentials, and conserver must be stop and started.
-d|--database
This option will reinitialize the basic xCAT database table setup. It will not remove any new database entries that have been added, but it is strongly suggested that you backup you database (dumpxCATdb) before using it.
-f|--force
The force option may be used after the install to reinitialize the Management Node. This option will regenerate keys, credential and reinititialize the site table. This option should be used, if keys or credentials become corrupt or lost. Additional action must be taken after using the force options. ssh keys must be redistributed to the nodes, site table attributes might need to be restored, makeconservercf needs to be rerun to pick up the new credentials and conserver stoped and started, rspconfig needs to be rerun to distribute the new keys to the MM and the HMCs. A new set of common ssh host keys will have been generated for the nodes. If you wish your nodes to be able to ssh to each other with out password intervention, then you should redistribute these new keys to the nodes. If the nodes hostkeys are updated then you will need to remove their entries from the known_hosts files on the management node before using ssh, xdsh, xdcp. Redistribute credentials and ssh keys to the service nodes and ssh keys to the nodes by using the updatenode -k command.
-m|--mgtnode
This option will add the Management Node to the database with the correct attributes set to be recognized by xCAT. This should be run after the hostname of the Management Node is set to the name that will resolve to the cluster-facing NIC.
-t|--tunables
This option will set tunable parameters on the Management and Service nodes recommended for your Linux cluster. It will only set them during initial install, if you run xcatconfig -f or xcatconfig -t.
*
To force regeneration of keys and credentials and reinitialize the site table:
xcatconfig -f
*
To regenerate root’s ssh keys:
xcatconfig -k
*
To regenerate node host ssh keys:
xcatconfig -s
*
To regenerate node host ssh keys and credentials:
xcatconfig -s -c
*
To add the Management Node to the DB:
xcatconfig -m
xcatd.8¶
xcatd - The xCAT daemon
xcatd
The heart of the xCAT architecture is the xCAT daemon xcatd on the management node. This receives requests from the client, validates the requests, and then invokes the operation. The xcatd daemon also receives status and inventory info from the nodes as they are being discovered and installed/booted.
Errors and information are reported through syslog to the /var/log/messages file. You can search for xCAT in those messages.
For further information: See https://sourceforge.net/apps/mediawiki/xcat/index.php?title=XCAT_2_Architecture.
To start/stop/restart xcatd on Linux, enter:
service xcatd start service xcatd stop service xcatd restart
To start/stop/restart xcatd on AIX, enter:
restartxcatd or startsrc -s xcatd stopsrc -s xcatd
/opt/xcat/sbin/xcatd
xcatdebug.8¶
xcatdebug - Enable or disable the trace facilities for xCAT. (Only supports Linux Operating System)
xcatdebug { [-f enable|disable [-c configuration file | subroutine list]] | [ -d enable |disable]}
xCAT offers two trace facilities to debug the xCAT:
*
Subroutine calling trace
Display the calling trace for subroutine when it is called.
The trace message includes: The name of the called subroutine; The arguments which passed to the called subroutine; The calling stack of the subroutine. By default, the trace will be enabled to all the subroutines in the xcatd and plugin modules. The target subroutine can be configured by configuration file or through xcatdebug command line.
- The flag -c is used to specify the subroutine list for subroutine calling trace, it can only work with -f. The value of -c can be a configuration file or a subroutine list.
- configuration file: a file contains multiple lines of SUBROUTINE_DEFINITIONsubroutine list: SUBROUTINE_DEFINITION | SUBROUTINE_DEFINITION|...
SUBROUTINE_DEFINITION: is the element for the -c to specify the subroutine list.
The format of SUBROUTINE_DEFINITION: [plugin](subroutine1,subroutine2,...)
- If ignoring the [plugin], the subroutines in the () should be defined in the xcatd.
- e.g. (daemonize,do_installm_service,do_udp_service)
- Otherwise, the package name of the plugin should be specified.
- e.g. xCAT::Utils(isMN,Version) e.g. xCAT_plugin::DBobjectdefs(defls,process_request)
The trace log will be written to /var/log/xcat/subcallingtrace. The log file subcallingtrace will be backed up for each running of the xcatdebug -f enable.
*
Commented trace log
The trace log code is presented as comments in the code of xCAT. In general mode, it will be kept as comments. But in debug mode, it will be commented back as common code to display the trace log.
NOTE: This facility can be enabled by pass the ENABLE_TRACE_CODE=1 global variable when running the xcatd. e.g. ENABLE_TRACE_CODE=1 xcatd -f
This facility offers two formats for the trace log code:
*
- Trace section
- ## TRACE_BEGIN # print “In the debugn”; ## TRACE_END
*
- Trace in a single line
- ## TRACE_LINE print “In the trace linen”;
The commented trace log can be added in xcatd and plugin modules. But following section has been added into the BEGIN {} section of the target plugin module to enable the facility.
if (defined $ENV{ENABLE_TRACE_CODE}) { use xCAT::Enabletrace qw(loadtrace filter); loadtrace(); }
-f
Enable or disable the subroutine calling trace.
For enable, if ignoring the -c flag, all the subroutines in the xcatd and plugin modules will be enabled.
For disable, all the subroutines which has been enabled by -f enable will be disabled. -c will be ignored.
-c
Specify the configuration file or subroutine list.
*
- configuration file\ : a file contains multiple lines of \ SUBROUTINE_DEFINITION
- e.g.
- (plugin_command) xCAT_plugin::DBobjectdefs(defls,process_request) xCAT::DBobjUtils(getobjdefs)
- * subroutine list\ : a string like \ SUBROUTINE_DEFINITION | SUBROUTINE_DEFINITION|...
- e.g.
- “(plugin_command)|xCAT_plugin::DBobjectdefs(defls,process_request)|xCAT::DBobjUtils(getobjdefs)”
-d
Enable or disable the commented trace log.
Note: The xcatd will be restarted for the performing of -d.
1
- Enable the subroutine calling trace for all the subroutines in the xcatd and plugin modules.
- xcatdebug -f enable
2
- Enable the subroutine calling trace for the subroutines configured in the /opt/xcat/share/xcat/samples/tracelevel0
- xcatdebug -f enable -c /opt/xcat/share/xcat/samples/tracelevel0
3
- Enable the subroutine calling trace for the plugin_command in xcatd and defls,process_request in the xCAT_plugin::DBobjectdefs module.
- xcatdebug -f enable -c “xCAT_plugin::DBobjectdefs(defls,process_request)|(plugin_command)”
4
- Disable the subroutine calling trace for all the subroutines which have been enabled by xcatdebug -f enable.
- xcatdebug -f disable
5
- Enable the commented trace log
- xcatdebug -d enable
6
- Enable both the subroutine calling trace and commented trace log
- xcatdebug -f enable -c /opt/xcat/share/xcat/samples/tracelevel0 -d enable
xcatsetup.8¶
xcatsetup - Prime the xCAT database using naming conventions specified in a config file.
xcatsetup [-s|--stanzas stanza-list] [--yesreallydeletenodes] cluster-config-file
xcatsetup [-? | -h | --help | -v | --version]
The xcatsetup command reads the specified config file that contains general information about the cluster being set up, and naming conventions and IP addresses that you want to use. It then defines the basic objects in the xCAT database representing this cluster configuration. The xcatsetup command prepares the database for the step of discovering the hardware that is connected to the service and cluster networks. The typical steps of setting up a system p cluster are:
*
Install the xCAT software on the management node
*
Create the cluster config file and run xcatsetup
*
Put hardware control passwords in the ppchcp or ppcdirect database table
*
Run makenetworks and makedhcp
*
Run the discovery commands (lsslp, mkhwconn, rspconfig) as described in the System P Hardware Management cookbook.
*
Configure and start services using makehosts, makedns, mkconserver.cf, etc.
*
Create the images that should be installed or booted on the nodes
*
Run nodeset and rpower/rnetboot to boot up the nodes.
The xcatsetup command is intended as a quick way to fill out the database for a cluster that has very regular naming patterns. The only thing it does is fill in database attributes. If your cluster does not follow consistent naming patterns, or has some other special configuration, you should define attribute values manually using mkdef(1)|mkdef.1, instead of using xcatsetup. The cluster config file is meant to be an easy way to prime the database; it is not meant to be a long living file that you update as the cluster changes. If you do want to run xcatsetup again at a later time, because, for example, you added a lot of nodes, you should put the total list of nodes in the config file, not just the new ones. This is because xcatsetup uses some regular expressions for groups (e.g. frame, cec, compute) that would be calculated incorrectly if the config file told xcatsetup about only the new nodes.
Speaking of regular expressions, xcatsetup creates some pretty complicated regular expressions in the database. These are useful because they keep most of the tables small, even for large clusters. But if you want to tweak them, they may be hard to understand. If after running xcatsetup, you want to convert your database to use individual rows for every node, you can do the following:
lsdef -z all >tmp.stanza
cat tmp.stanza | chdef -z
Many of the sections and attributes in the configuration file can be omitted, if you have a simple cluster, or if you want to create just 1 or 2 of the object types at this time. See the section A Simpler Configuration File for an example of this.
If you want to delete all of the nodes that xcatsetup created, and start over, use the --yesreallydeletenodes option.
*
The xcatsetup command has only been implemented and tested for system p servers so far.
The config file is organized in stanza format and supports the keywords in the sample file below. Comment lines begin with “#”. Stanzas can be ommitted if you do not want to define that type of object. The only hostname formats supported are those shown in this sample file, although you can change the base text and the numbers. For example, hmc1-hmc3 could be changed to hwmgmt01-hwmgmt12. The hostnames specified must sort correctly. I.e. use node01-node80, instead of node1-node80. This sample configuration file is for a 2 building block cluster.
xcat-site:
domain = cluster.com
# currently only direct fsp control is supported
use-direct-fsp-control = 1
# ISR network topology. For example, one of the following: 128D, 64D, 32D, 16D, 8D, 4D, 2D, 1D
topology = 32D
# The nameservers in site table will be set with the value of master automatically.
xcat-service-lan:
# IP range used for DHCP. If you set the entry, the networks table will be filled
# automatically with this range and the dhcp interface will be set in the site table.
dhcp-dynamic-range = 50.0.0.0-50.0.0.200
xcat-hmcs:
hostname-range = hmc1-hmc2
starting-ip = 10.200.1.1
xcat-frames:
# these are the connections to the frames
hostname-range = frame[1-6]
num-frames-per-hmc = 3
# this lists which serial numbers go with which frame numbers
vpd-file = vpd2bb.stanza
# There are two rules of defining FSP/BPAs. The first defining the node's host name by increasing the last bit
# of IP address, while the second defining the node's name by varying the second bit and the third bit of IP.
# This assumes you have 2 service LANs: a primary service LAN 10.230.0.0/255.255.0.0 that all of the port 0's
# are connected to, and a backup service LAN 10.231.0.0/255.255.0.0 that all of the port 1's are connected to.
# bpa-a-0-starting-ip = 10.230.1.1
# bpa-b-0-starting-ip = 10.230.2.1
# bpa-a-1-starting-ip = 10.231.1.1
# bpa-b-1-starting-ip = 10.231.2.1
# This assumes you have 2 service LANs: a primary service LAN 40.x.y.z/255.0.0.0 that all of the port 0's
# are connected to, and a backup service LAN 41.x.y.z/255.0.0.0 that all of the port 1's are connected to.
# "x" is the frame number and "z" is the bpa/fsp id (1 for the first BPA/FSP in the Frame/CEC, 2 for the
# second BPA/FSP in the Frame/CEC). For BPAs "y" is always be 0 and for FSPs "y" is the cec id.
vlan-1 = 40
vlan-2 = 41
xcat-cecs:
# These are the connections to the CECs. Either form of hostname is supported.
#hostname-range = cec01-cec64
hostname-range = f[1-6]c[01-12]
# If you use the frame/cec hostname scheme above, but do not have a consistent
# number of cecs in each frame, xcat can delete the cecs that do not get
# supernode numbers assigned to them.
delete-unused-cecs = 1
# lists the HFI supernode numbers for each group of cecs in each frame
supernode-list = supernodelist2bb.txt
# If you do not want to specify the supernode-list at this time and you have a consistent
# number of cecs in each frame, you can instead just use this setting:
num-cecs-per-frame = 12
#fsp-a-0-starting-ip = 10.230.3.1
#fsp-b-0-starting-ip = 10.230.4.1
#fsp-a-1-starting-ip = 10.231.3.1
#fsp-b-1-starting-ip = 10.231.4.1
xcat-building-blocks:
num-frames-per-bb = 3
num-cecs-per-bb = 32
xcat-lpars:
num-lpars-per-cec = 8
# If you set these, then do not set the corresponding attributes in the other node stanzas below.
# Except you still need to set xcat-service-nodes:starting-ip (which is the ethernet adapter)
#hostname-range = f[1-6]c[01-12]p[1-8]
hostname-range = f[1-6]c[01-12]p[01,05,09,13,17,21,25,29]
starting-ip = 10.1.1.1
aliases = -hf0
# ml0 is for aix. For linux, use bond0 instead.
otherinterfaces = -hf1:11.1.1.1,-hf2:12.1.1.1,-hf3:13.1.1.1,-ml0:14.1.1.1
xcat-service-nodes:
num-service-nodes-per-bb = 2
# which cecs within the bldg block that the SNs are located in
cec-positions-in-bb = 1,32
# this is for the ethernet NIC on each SN
#hostname-range = sn1-sn4
starting-ip = 10.10.1.1
# this value is the same format as the hosts.otherinterfaces attribute except
# the IP addresses are starting IP addresses
#otherinterfaces = -hf0:10.10.1.1,-hf1:10.11.1.1,-hf2:10.12.1.1,-hf3:10.13.1.1,-ml0:10.14.1.1
xcat-storage-nodes:
num-storage-nodes-per-bb = 3
# which cecs within the bldg block that the storage nodes are located in
cec-positions-in-bb = 12,20,31
#hostname-range = stor1-stor6
#starting-ip = 10.20.1.1
#aliases = -hf0
#otherinterfaces = -hf1:10.21.1.1,-hf2:10.22.1.1,-hf3:10.23.1.1,-ml0:10.24.1.1
xcat-compute-nodes:
#hostname-range = n001-n502
#starting-ip = 10.30.1.1
#aliases = -hf0
# ml0 is for aix. For linux, use bond0 instead.
#otherinterfaces = -hf1:10.31.1.1,-hf2:10.32.1.1,-hf3:10.33.1.1,-ml0:10.34.1.1
The vpd-file specifies the following vpd table attributes for the frames: node, serial, mtm, side. Use the same stanza format that accepted by the chdef(1)|chdef.1 command, as documented in xcatstanzafile(5)|xcatstanzafile.5. The purpose of this file is to enable xCAT to match up frames found through lsslp(1)|lsslp.1 discovery with the database objects created by xcatsetup. All of the frames in the cluster must be specified.
Here is a sample file:
frame1:
objtype=node
serial=99200G1
mtm=9A00-100
frame2:
objtype=node
serial=99200D1
mtm=9A00-100
frame3:
objtype=node
serial=99200G1
mtm=9A00-100
frame4:
objtype=node
serial=99200D1
mtm=9A00-100
frame5:
objtype=node
serial=99200G1
mtm=9A00-100
frame6:
objtype=node
serial=99200D1
mtm=9A00-100
The supernode-list file lists what supernode numbers should be given to each CEC in each frame. Here is a sample file:
frame1: 0, 1, 16
frame2: 17, 32
frame3: 33, 48, 49
frame4: 64 , 65, 80
frame5: 81, 96
frame6: 97(1), 112(1), 113(1), 37(1), 55, 71
The name before the colon is the node name of the frame. The numbers after the colon are the supernode numbers to assign to the groups of CECs in that frame from bottom to top. Each supernode contains 4 CECs, unless it is immediately followed by “(#)”, in which case the number in parenthesis indicates how many CECs are in this supernode.
This is an example of a simple cluster config file that just defines the frames and CECs for 2 frames, without specifying VPD data or supernode numbers at this time.
xcat-site:
use-direct-fsp-control = 1
xcat-frames:
hostname-range = frame[1-2]
xcat-cecs:
#hostname-range = cec[01-24]
hostname-range = f[1-2]c[01-12]
num-cecs-per-frame = 12
xcat-lpars:
hostname-range = f[1-2]c[01-12]p[01,05,09,13,17,21,25,29]
The following lists which database attributes are filled in as a result of each stanza. Note that depending on the values in the stanza, some attributes might not be filled in.
xcat-site
site table: domain, nameservers, topology
xcat-hmcs
site table: ea_primary_hmc, ea_backup_hmc
nodelist table: node, groups (all HMCs (hmc) ), hidden
hosts table: node, ip
ppc table: node, comments
nodetype table: node, nodetype
xcat-frames
nodelist table: node, groups (all frames (frame) ), hidden
ppc table: node, id, hcp, nodetype, sfp
nodetype table: node, nodetype
nodehm table: node, mgt
vpd table: node, serial, mtm, side
xcat-bpas
nodelist table: node, groups (bpa,all) , hidden
ppc table: node, id, hcp, nodetype, parent
nodetype table: node, nodetype
nodehm table: node, mgt
vpd table: node, serial, mtm, side
xcat-cecs
nodelist table: node, groups (all CECs (cec), all CECs in a frame (<frame>cec) ), hidden
ppc table: node, supernode, hcp, id, parent
nodetype table: node, nodetype
nodehm table: node, mgt
nodegroup table: groupname, grouptype, members, wherevals (all nodes in a CEC (<cec>nodes) )
nodepos: rack, u
xcat-fsps
nodelist table: node, groups (fsp,all), hidden
ppc table: node, id, hcp, nodetype, parent
nodetype table: node, nodetype
nodehm table: node, mgt
vpd table: node, serial, mtm, side
xcat-building-blocks
site table: sharedtftp, sshbetweennodes(service)
ppc table: node, parent (for frame)
xcat-service-nodes
nodelist table: node, groups (all service nodes (service), all service nodes in a BB (bb<num>service) )
hosts table: node, ip, hostnames, otherinterfaces
ppc table: node, id, hcp, parent
nodetype table: node, nodetype, arch
nodehm table: node, mgt, cons
noderes table: netboot
servicenode table: node, nameserver, dhcpserver, tftpserver, nfsserver, conserver, monserver, ftpserver, nimserver, ipforward
nodegroup table: groupname, grouptype, members, wherevals (all nodes under a service node (<servicenode>nodes) )
nodepos: rack, u
xcat-storage-nodes
nodelist table: node, groups (all storage nodes (storage), all storage nodes in a BB (bb<num>storage) )
hosts table: node, ip, hostnames, otherinterfaces
ppc table: node, id, hcp, parent
nodetype table: node, nodetype, arch
nodehm table: node, mgt, cons
noderes table: netboot, xcatmaster, servicenode
nodepos: rack, u
xcat-compute-nodes
nodelist table: node, groups (all compute nodes (compute) )
hosts table: node, ip, hostnames, otherinterfaces
ppc table: node, id, hcp, parent
nodetype table: node, nodetype, arch
nodehm table: node, mgt, cons
noderes table: netboot, xcatmaster, servicenode
nodepos: rack, u
ll-config
postscripts: postscripts
-s|--stanzas stanza-list
A comma-separated list of stanza names that xcatsetup should process in the configuration file. If not specified, it will process all the stanzas that start with ‘xcat’ and some other stanzas that give xCAT hints about how to set up the HPC products.
This option should only be specified if you have already run xcatsetup earlier with the stanzas that occur before this in the configuration file. Otherwise, objects will be created that refer back to other objects that do not exist in the database.
-v|--version
Command Version.
-?|-h|--help
Display usage message.
--yesreallydeletenodes
Delete the nodes represented in the cluster config file, instead of creating them. This is useful if your first attempt with the cluster config file wasn’t quite right and you want to start over. But use this option with extreme caution, because it will potentially delete a lot of nodes. If the only thing you have done so far in your database is add nodes by running xcatsetup, then it is safe to use this option to start over. If you have made other changes to your database, you should first back it up using dumpxCATdb(1)|dumpxCATdb.1 before using this option.
Use the sample config.txt file at the beginning of this man page to create all the objects/nodes for a 2 building block cluster.
xcatsetup config.txt
The output:
Defining site attributes... Defining HMCs... Defining frames... Defining CECs... Defining building blocks... Defining LPAR nodes...
Use the simpler config file shown earlier in this man page to create just the frame and cec objects:
xcatsetup config-simple.txt
The output:
Defining frames... Defining CECs...
/opt/xcat/sbin/xcatsetup
mkdef(1)|mkdef.1, chdef(1)|chdef.1, lsdef(1)|lsdef.1, xcatstanzafile(5)|xcatstanzafile.5, noderange(3)|noderange.3, nodeadd(8)|nodeadd.8
xcatsnap.8¶
xcatsnap - Gathers information for service about the current running xCAT environment.
xcatsnap {-h | --help}
xcatsnap {-v | --version}
xcatsnap {-B | --BYPASS}
xcatsnap {-d | --dir}
xcatsnap - The xcatsnap command gathers configuration, log and trace information about the xCAT components that are installed. This command only collects the data on the local node on which this command is run. This command is typically executed when a problem is encountered with any of these components in order to provide service information to the IBM Support Center.
This command should only be executed at the instruction of the IBM Support Center.
-h|--help
Displays the usage message.
-v|--version
Displays the release version of the code.
-B|--bypass
Runs in bypass mode, use if the xcatd daemon is hung.
-d|--dir
The directory to put the snap information. Default is /tmp/xcatsnap.
*
Run the xcatsnap routine in bypass mode and put info in /tmp/mydir :
xcatsnap -B -d /tmp/mydir
*
To run the xcatsnap routine and use default directory /tmp/xcatsnap :
xcatsnap
Security Notices¶
2016 Notices¶
2016-11-30 - Removal of Service Stream Password¶
It has been brought to our attention that the xCAT product has hard-coded default passwords for the HMC/FSP to allow for IBM Service to connect to customer machines for L2/L3 support activities. This creates a security vulnerability where third parties could potentially gain root level access using these weak, hard coded passwords.
Example:
create_pwd => "netsDynPwdTool --create dev FipSdev", password => "FipSdev"
In response, xCAT will remove these hard-coded password and interfaces from the xCAT code.
Action¶
No action is required for xCAT 2.12.3, and higher.
If running older versions of xCAT, update xCAT to a higher level code base that has the hard-coded default passwords removed.
The following table describes the recommended update path:
| xCAT Version | Action | Release Notes |
|---|---|---|
| 2.13, or newer | No applicable | |
| 2.12.x | Update to 2.12.3, or higher | 2.12.3 Release Notes |
| 2.11.x | Update to 2.12.3, or higher | 2.12.3 Release Notes |
| 2.10.x | Update to 2.12.3, or higher | 2.12.3 Release Notes |
| 2.9.x, or older | Update to:
|
2.9.4 Release Notes |