What were you trying to do that didn't work?

db2 resource agent use flow for incident
 

Incident (Failover/Migration)

• Trigger: Monitor failure on Primary.

• Sequence of Timeouts:

             Demote (): Pacemaker tries db2 deactivate db.

             Escalation to Stop (): Once the timeout expires, Pacemaker kills the demote action and issues a stop.

             Fencing: If the stop fails after stop-timeout it will run db2nkill then (or completes), the node is fenced or the resource moves.

IBM opinion on the topic :

"db2stop force" - which does a deactivate which in turn drives log truncation and can cause the log streams to diverge - so in any automation driven event it would actually always be safer just to kill  e.g. through pkill or db2gcf -k.

What is the impact of this issue to you?

divergence in the log will cause inconsistency and customer will have to rebuild DB

Please provide the package NVR for which the bug is seen:

How reproducible is this bug?:

Steps to reproduce

Expected results

Actual results