PowerHA SystemMirror fallover is not triggered after a node crash or reboot

Problem

PowerHA® SystemMirror® fails to selectively move the affected resource group to another cluster node when a node crashes or restarts.

Solution

Check if either more than half of the nodes in the cluster are online or exactly half of the nodes are online and the tiebreaker is reserved. If less than half of the nodes are online, start additional nodes. If exactly half of the nodes are online, check the attribute of the active tiebreaker.

The possible causes follow:
  • Either more than half of the nodes in the cluster are online or exactly half of the nodes are online and the tiebreaker is reserved. If less than half of the nodes are online, start additional nodes. If exactly half of the nodes are online, check the attribute of the active tiebreaker by using the clmgr query cluster command.
  • The split policy must be set as None and if it is set as Manual, the user intervention is required for fallover after node reboots.
  • The policy of resource group must be FBHN, which can be checked by using the clmgr query rg command.