Chapter 4. Clustering

systemd and pacemaker now coordinate correctly during system shutdown

Previously, systemd and pacemaker did not coordinate correctly during system shutdown, which caused pacemaker resources not to be terminated properly. With this update, pacemaker is ordered to stop before dbus and other systemd services that pacemaker started. This allows both pacemaker and the resources that pacemaker manages to shut down properly.

The pcs resource move and pcs resource ban commands now display a warning message to clarify the commands' behavior

The pcs resource move command and the pcs resource ban commands create location constraints that that effectively ban the resource from running on the current node until the constraint is removed or until the constraint lifetime expires. This behavior had previously not been clear to users. These commands now display a warning message explaining this behavior, and the help screens and documentation for these commands have been clarified.

New command to move a Pacemaker resource to its preferred node

After a Pacemaker resource has moved, either due to a failover or to an administrator manually moving the node, it will not necessarily move back to its original node even after the circumstances that caused the failover have been corrected. You can now use the pcs resource relocate run command to move a resource to its preferred node, as determined by current cluster status, constraints, location of resources and other settings. You can also use the pcs resource relocate show command to display migrated resources. For information on these commands, see the High Availability Add-On Reference.

Simplified method for configuring fencing for redundant power supplies in a cluster

When configuring fencing for redundant power supplies, you must ensure that when the power supplies are rebooted both power supplies are turned off before either power supply is turned back on. If the node never completely loses power, the node may not release its resources. This opens up the possibility of nodes accessing these resources simultaneously and corrupting them.
Prior to Red Hat Enterprise Linux 7.2, you needed to explicitly configure different versions of the devices which used either the 'on' or 'off' actions. Since Red Hat Enterprise Linux 7.2, it is now only required to define each device once and to specify that both are required to fence the node.
For information on configuring fencing for redundant power supplies, see the Fencing: Configuring STONITH chapter of the High Availability Add-On Reference manual.

New --port-as-ip option for fencing agents

Fence agents used only with single devices required complex configuration in pacemaker. It is now possible to use the --port-as-ip option to enter the IP address in the port option.