tomcat resource frequently timing out on start and showing as stopped in pcs output in a RHEL 7 High Availability cluster with pacemaker

Issue

• lrmd shows timeouts for a tomcat resource during start operations, and pcs reports it as stopped

Oct 21 05:35:33 node1 lrmd[1769]: warning: child_timeout_callback: tomcat_start_0 process (PID 459) timed out
Oct 21 05:35:33 node1 lrmd[1769]: warning: operation_finished: tomcat_start_0:459 - timed out after 60000ms
Oct 21 05:35:33 node1 crmd[1772]: error: process_lrm_event: LRM operation tomcat_start_0 (222) Timed Out (timeout=60000ms)

• pcs shows a "Failed action" for a tomcat resource with "status=Timed Out"

Failed actions:
    tomcat_start_0 on node1.example.com 'unknown error' (1): call=222, status=Timed Out, last-rc-change='Tue Oct 21 05:34:33 2014', queued=60002ms, exec=0ms

• Periodically, tomcat seems to disassociate itself from pacemaker. tomcat is seen as running in ps output but it shows stopped in pcs status. Sometimes, tomcat completely fails.