Useful Corosync and Pacemaker Commands

List of useful commands for Corosync and Pacemaker clusters.

• View the real-time status of nodes and resources:

crm_mon

• View one-off status of nodes and resources:

crm status

• View cluster configuration:

crm configure show

• Set a node to standby mode (can also be used to simulate a node becoming unavailable):

crm node standby <node name>

• Set a node's status from standby to online:

crm node online <node name>

• Start a resource or a resource group:

crm resource start <resource name>

• Stop a resource or a resource group:

crm resource stop <resource name>

• Restart a resource or a resource group (if the resource is in a resource group then the whole group will be restarted):

crm resource restart <resource name>

• Edit a resource, which allows its reconfiguration:

crm configure edit <resource name>

• Delete a resource (must be stopped before it is deleted):

crm configure delete <resource name>

• Migrate the enswitch resource group to another node (avoid migrating specific resources separately, migrate the whole enswitch group instead):

crm resource migrate enswitch <node name>

• Configuring a cluster to not manage a resource:

crm_resource --resource <resource name> --set-parameter is-managed --meta --parameter-value false

• Clear a resource that is stuck in "(unmanaged) FAILED" state, e.g. due to a failed stop action:

crm_resource -P
crm resource cleanup <resource name>

• List resource agents by type:

crm ra list lsb
crm ra list systemd
crm ra list ocf heartbeat
crm ra list ocf pacemaker

• Remove a resource from a group without disturbing the remaining resources:

crm configure property maintenance-mode=true
crm resource stop <resource> # it won't stop as it's in maintenance-mode
crm configure delete <resource>
crm configure show # verify that all references to <resource> are gone
crm resource reprobe # the cluster double checks the status of declared resources and verifies that everything is in order and <resource> doesn't exist anymore
crm_mon -Arf1 # double check that everything is "started (unmanaged)" and <resource> is gone
crm_simulate -S -L -VVV # optional, to check what would happen when leaving maintenance-mode
crm configure property maintenance-mode=false

If something goes wrong while in maintenance-mode then

crm resource cleanup <resource>

might be handy.

• Edit the cluster configuration with the cibadmin cluster tool:

cibadmin --query > /tmp/old_cluster_configuration.xml
cp /tmp/old_cluster_configuration.xml /tmp/new_cluster_configuration.xml
# with any editor, e.g vim, apply any needed changes to the cluster configuration
vim /tmp/new_cluster_configuration.xml
cibadmin --replace --xml-file /tmp/new_cluster_configuration.xml

That sequence of commands needs to be executed in only one cluster node. After that, all the other nodes will receive and apply the updated configuration.

• Set custom values for the monitoring interval and stop timeout for a systemd resource:

In the following example a cluster resource of systemd type will be created, setting a monitoring interval of 30 seconds and a stop timeout of 40 seconds.

crm configure primitive <resource name> \
  systemd:<sytemd name that corresponds to the cluster resource> \
  op monitor interval="30s" \
  op stop timeout="40s"

If the cluster resource already exists, it could be edited to set those custom values, as illustrated in the previous examples.