Search code examples
cluster-computingpacemaker

pacemaker corosync service ignored

Two Node cluster Node A , Node B .

Service X running on Node A, Node B is DC.

We are using stack corosync with Pacemaker. Failure Timeout is 10 sec . Target-Role is started .

Events happens like this Node A sends event to Node B Service X is down Node B prints Ignoring expired failure for Service X After this Service X is never restarted by the Cluster.

Now questions are:

  1. Why is Node B (DC) ignoring the expired failure?
  2. Even for this time DC ignored but as the Service X is down, Node A should monitor the service and again send failure status to Node B and at that time Node B should restart the service. Why this no hapenning?
Solution

  • One Reason for this may be time difference between two servers (DC and Other Machine) .

    So , DC thinks that this event is old and ignore it . Please sync time and then try to re-create the issue .