Automated monitoring with Pacemaker resource agents

Being There


When a cluster node fails, the Pacemaker high-availability tool launches the services on another node. A lesser known feature is Pacemaker’s ability to put failed services back on their feet in the cluster manager.

Pacemaker [1] has developed into the undisputed, standard high-availability solution for Linux systems. If one of two nodes in a cluster fails, Pacemaker launches the failed services on the other cluster node. But if you only use Pacemaker for failover actions, you are missing one of its best features. Many administrators are unaware that Pacemaker can automatically put failed applications back on their feet.