Chapter 3. The pcs command-line interface

The pcs command-line interface controls and configures cluster services such as corosync, pacemaker,booth, and sbd by providing an easier interface to their configuration files.

Note that you should not edit the cib.xml configuration file directly. In most cases, Pacemaker will reject a directly modified cib.xml file.

3.1. pcs help display

You use the -h option of pcs to display the parameters of a pcs command and a description of those parameters.

The following command displays the parameters of the pcs resource command.

# pcs resource -h

3.2. Viewing the raw cluster configuration

Although you should not edit the cluster configuration file directly, you can view the raw cluster configuration with the pcs cluster cib command.

You can save the raw cluster configuration to a specified file with the pcs cluster cib filename command. If you have previously configured a cluster and there is already an active CIB, you use the following command to save the raw xml file.

pcs cluster cib filename

For example, the following command saves the raw xml from the CIB into a file named testfile.

# pcs cluster cib testfile

3.3. Saving a configuration change to a working file

When configuring a cluster, you can save configuration changes to a specified file without affecting the active CIB. This allows you to specify configuration updates without immediately updating the currently running cluster configuration with each individual update.

For information about saving the CIB to a file, see Viewing the raw cluster configuration. Once you have created that file, you can save configuration changes to that file rather than to the active CIB by using the -f option of the pcs command. When you have completed the changes and are ready to update the active CIB file, you can push those file updates with the pcs cluster cib-push command.

Procedure

The following is the recommended procedure for pushing changes to the CIB file. This procedure creates a copy of the original saved CIB file and makes changes to that copy. When pushing those changes to the active CIB, this procedure specifies the diff-against option of the pcs cluster cib-push command so that only the changes between the original file and the updated file are pushed to the CIB. This allows users to make changes in parallel that do not overwrite each other, and it reduces the load on Pacemaker which does not need to parse the entire configuration file.

  1. Save the active CIB to a file. This example saves the CIB to a file named original.xml.

    # pcs cluster cib original.xml
  2. Copy the saved file to the working file you will be using for the configuration updates.

    # cp original.xml updated.xml
  3. Update your configuration as needed. The following command creates a resource in the file updated.xml but does not add that resource to the currently running cluster configuration.

    # pcs -f updated.xml resource create VirtualIP ocf:heartbeat:IPaddr2 ip=192.168.0.120 op monitor interval=30s
  4. Push the updated file to the active CIB, specifying that you are pushing only the changes you have made to the original file.

    # pcs cluster cib-push updated.xml diff-against=original.xml

Alternately, you can push the entire current content of a CIB file with the following command.

pcs cluster cib-push filename

When pushing the entire CIB file, Pacemaker checks the version and does not allow you to push a CIB file which is older than the one already in a cluster. If you need to update the entire CIB file with a version that is older than the one currently in the cluster, you can use the --config option of the pcs cluster cib-push command.

pcs cluster cib-push --config filename

3.4. Displaying cluster status

There are a variety of commands you can use to display the status of a cluster and its components.

You can display the status of the cluster and the cluster resources with the following command.

# pcs status

You can display the status of a particular cluster component with the commands parameter of the pcs status command, specifying resources, cluster, nodes, or pcsd.

pcs status commands

For example, the following command displays the status of the cluster resources.

# pcs status resources

The following command displays the status of the cluster, but not the cluster resources.

# pcs cluster status

3.5. Displaying the full cluster configuration

Use the following command to display the full current cluster configuration.

# pcs config

3.6. Modifying the corosync.conf file with the pcs command

You can use the pcs command to modify the parameters in the corosync.conf file.

The following command modifies the parameters in the corosync.conf file.

pcs cluster config update [transport pass:quotes[transport options]] [compression pass:quotes[compression options]] [crypto pass:quotes[crypto options]] [totem pass:quotes[totem options]] [--corosync_conf pass:quotes[path]]

The following example command udates the knet_pmtud_interval transport value and the token and join totem values.

# pcs cluster config update transport knet_pmtud_interval=35 totem token=10000 join=100

Additional resources

3.7. Displaying the corosync.conf file with the pcs command

The following command displays the contents of the corosync.conf cluster configuration file.

# pcs cluster corosync

You can print the contents of the corosync.conf file in a human-readable format with the pcs cluster config command, as in the following example.

The output for this command includes the UUID for the cluster if the cluster was created in RHEL 9.1 or later, or if the UUID was added manually as described in Identifying clusters by UUID.

[root@r8-node-01 ~]# pcs cluster config
Cluster Name: HACluster
Cluster UUID: ad4ae07dcafe4066b01f1cc9391f54f5
Transport: knet
Nodes:
  r8-node-01:
    Link 0 address: r8-node-01
    Link 1 address: 192.168.122.121
    nodeid: 1
  r8-node-02:
    Link 0 address: r8-node-02
    Link 1 address: 192.168.122.122
    nodeid: 2
Links:
  Link 1:
    linknumber: 1
    ping_interval: 1000
    ping_timeout: 2000
    pong_count: 5
Compression Options:
  level: 9
  model: zlib
  threshold: 150
Crypto Options:
  cipher: aes256
  hash: sha256
Totem Options:
  downcheck: 2000
  join: 50
  token: 10000
Quorum Device: net
  Options:
    sync_timeout: 2000
    timeout: 3000
  Model Options:
    algorithm: lms
    host: r8-node-03
  Heuristics:
    exec_ping: ping -c 1 127.0.0.1

You can run the pcs cluster config show command with the --output-format=cmd option to display the pcs configuration commands that can be used to recreate the existing corosync.conf file, as in the following example.

[root@r8-node-01 ~]# pcs cluster config show --output-format=cmd
pcs cluster setup HACluster \
  r8-node-01 addr=r8-node-01 addr=192.168.122.121 \
  r8-node-02 addr=r8-node-02 addr=192.168.122.122 \
  transport \
  knet \
    link \
      linknumber=1 \
      ping_interval=1000 \
      ping_timeout=2000 \
      pong_count=5 \
    compression \
      level=9 \
      model=zlib \
      threshold=150 \
    crypto \
      cipher=aes256 \
      hash=sha256 \
  totem \
    downcheck=2000 \
    join=50 \
    token=10000