 |
» |
|
|
|
This section describes the HP XC command environment. The HP XC system commands were developed to supplement the Linux command set and other Open Source commands. HP XC Command Set | |
Table 1-1 lists the commands alphabetically and provides a brief description. For more information on an individual command, see the corresponding manpage. Table 1-1 HP XC System Commands | Command | Description |
|---|
cexec | The cexec command is a shell script that invokes the pdsh command to perform commands on multiple node in the system. Manpage: cexec(1) | cluster_config | The cluster_config command enables you to view and modify the default role assignments and node configuration, modify the default role assignments on any node, and add, modify, or delete Ethernet connections to any node except the head node. When you are satisfied with the current configuration, the cluster_config utility populates the database with the assigned services for nodes, performs a global configuration of services on all nodes, runs a utility to create the golden system image, and runs node-specific cluster configuration scripts on the head node. | collectl | The collectl utility collects data on the nodes of the HP XC system and plays back the information as ASCII text or in a plot form. For more information, see The collectl Utility. Manpage: collectl(1) | console | The console command enables access to the consoles of all application nodes. Manpage: console(8) | controllsf | Use the controllsf command to control the execution of LSF-HPC on the HP XC system. Manpage: controllsf(8) | dbsysparams | Use the dbsysparams command to return or set the value of an attribute of the hptc_system table in the configuration and management database (cmdb). Manpage: dbsysparams(8) | headnode | This command returns the name of the head node. Manpage: headnode(1) | hostgroup | Use the hostgroup command to create, modify, or delete predefined named lists of nodes, called host groups. Manpage: hostgroup(1) | locatenode | Use this command to locate one or more nodes. Invoking this command with the --on option illuminates the Unit Identifier LED on the node's front panel (not all nodes have this feature). Manpage: locatenode(8) | managedb | Use the managedb command to perform an archive, a backup, or a dump of the configuration database. An archive function copies data from the configuration database and removes archived data. A backup of the configuration database copies the table data without changing the database. A dump prints the configuration database tables in an orderly format. Manpage: managedb(8) | nodename | This command displays the name of the node on which it is run. Manpage: nodename(1) | openipport | The superuser uses the openipport command to open a specified port in the firewall. Manpage: openipport(8) | ovp | Use the ovp utility to verify the installation, configuration, and operation of the HP XC system. Manpage: ovp(8) | power | Use the power command to control the power for a set of nodes and to interrogate their current state. You can also use it to turn an identifier light on or off. Manpage: power(8) | sacct | Use this command to display accounting data for all jobs and job steps in the SLURM job accounting log. Manpage: sacct(1) | setnode | Use this command to set the node settings in the configuration and management database. With this command, you can enable or disable nodes, or cause a re-imaging of specified nodes. Manpage: setnode(8) | shownode | This command has a variety of subcommands that provide information about nodes. Manpage: shownode(1) | ssh_create_shared_keys | The ssh_create_shared_keys command, used on a one-time basis, updates the appropriate ssh key files in the user's $HOME/.ssh directory so they do not need to provide a login name and password each time they log in to another node in the HP XC system or run jobs on those nodes. | startsys | The startsys command causes specific actions to take place to start the entire HP XC system or a subset of nodes. Manpage: startsys(8) | stopsys | The stopsys command causes specific actions to take place to halt the HP XC system or on a subset of nodes. Manpage: stopsys(8) | sys_check | The sys_check utility is a data collection tool that helps diagnose system errors and problems. It creates an HTML report of your system's configuration (hardware and software). Manpage: sys_check(8) | updateimage | Use this command to update the golden image on the image server based on the changes found on the golden master. When client nodes request an updated golden image, they receive the latest software in that image. Manpage: updateimage(8) |
Interpreting the nodelist Parameter | |
The nodelist parameter, used in several HP XC system commands, indicates one or more nodes. You can use brackets, hyphens, and commas in the nodelist parameter: - [ ]
Brackets indicate a set of nodes. You can use only one set of brackets for each instance of the nodelist parameter. - -
A hyphen indicates a range of nodes. - ,
A comma separates node definitions.
For example, the nodes n1 n2 n3 n5 can be expressed in the nodelist parameter as n[1-3,5]. Executing a Command on Multiple Nodes | |
You can enter the HP XC system commands at the command line or invoke them remotely on a node in the HP XC system using the pdsh shell or the cexec command. These commands enable you to execute a command on multiple nodes at the same time. Use the cexec command when you need to log the command. Use the pdsh command when logging is not required. Some commands, like the power command, accept a list of nodes as a parameter; this parameter is often referred to as a nodelist. These commands affect the nodes directly, without the need for the pdsh or cexec commands. You can use the HP XC commands in a cron script like any other command. The Open Source Parallel Distributed Shell (pdsh) command is a multithreaded remote shell client that executes commands on multiple nodes in parallel. You can specify all nodes, a given number of nodes, or only certain nodes to perform the command or commands passed as arguments to the pdsh command. By default, the pdsh shell can issue 32 simultaneous remote commands at a time. You can override this default with the -f option. The pdsh shell relies on the ssh transport mechanism. Ensure that ssh is properly installed on the system; changes to the ssh configuration might cause the pdsh shell to fail. Although the pdsh shell is capable of using several remote shell services, including rsh and ssh, the security settings of the HP XC system make ssh the shell of choice. The format of this command is: pdsh -[options] " command ... " |
The following example runs the uptime command on all the nodes in a four-node system. # pdsh -a "uptime"
n4: 15:51:40 up 2 days, 2:41, 4 users, load average: 0.48, 0.29, 0.11
n3: 15:49:17 up 1 day, 4:55, 0 users, load average: 0.00, 0.00, 0.00
n2: 15:50:32 up 1 day, 4:55, 0 users, load average: 0.00, 0.00, 0.00
n1: 15:47:21 up 1 day, 4:27, 0 users, load average: 0.00, 0.00, 0.00
|
For additional information, see pdsh(1). You can also find additional information at the following Web site: http://www.llnl.gov/linux/pdsh/ The cexec command provides the same facility as the pdsh command plus two additional features: The cexec command provides an additional option, which enables you to specify a host group or a service group: A host group is a designated list of nodes. A service group is a list of nodes and the services that are configured to run on them.
The output of the cexec command is logged as an event, but the output of the pdsh command is not.
The format of this command is: cexec -[options] " command ... " |
Useful command options include the following: The following example runs the users command on all the nodes in a four-node system: # cexec -a "users"
n4: root db dal guest
n3: cjg
n2: rc wk
n1: root tcr
|
The following example runs the who --count command on nodes n12 and n25: # cexec -a "who --count"
n12: root bg rmk
n12: # users=3
n25: root wra guest spg
n25: # users=4
|
For additional information, see cexec(1). |
Printable version
