LVM resource with exclusive=true fails with "timed out after 30000ms" in a RHEL High Availability cluster with pacemaker

Issue

• An LVM resource's monitor operation is frequently timing out whenever we issue a large copy to a file system
• My LVM resource keeps failing its monitor
• A resource group failed over to another node after the LVM resource failed

• pacemaker HA-LVM agent monitor operation failed

  Jun  4 14:17:47 node1 lrmd[6514]:  warning: child_timeout_callback: myVG_monitor_10000 process (PID 40301) timed out
  Jun  4 14:17:47 node1 lrmd[6514]:  warning: operation_finished: myVG_monitor_10000:40301 - timed out after 30000ms
  Jun  4 14:17:47 node1 crmd[6517]:    error: process_lrm_event: Operation myVG_monitor_10000: Timed Out (node=node1.example.com call=141, timeout=30000ms)
  

• The LVM resource is created using multipath devices, but a fluctuation in few sub paths to the multipath device causes pacemaker LVM monitor time out immediately. The multipath device still had half of the sub paths active, so pacemaker LVM monitor should not have got failed or timed out.