pacemaker not restarting mysql resource when Filesystem resource it depends on fails a status check in a RHEL 6 or 7 High Availability cluster

Issue

• Pacemaker not restarting mysql resource when Filesystem resource fails
• When a Filesystem resource fails a status check, the mysql resource that uses that file system doesn't ever get restarted. pengine logs that the resource needs a restart, but there seems to be a failure on stop because the file system is not available, and then when its started back up it says its already running

Sep 02 09:24:39 node1 pengine[3459]: notice: LogActions: Recover db_fs        (Started node1)
Sep 02 09:24:39 node1 pengine[3459]: notice: LogActions: Restart db_srv        (Started node1)
[...]
Sep 02 09:24:39 node1 crmd[3461]: notice: te_rsc_command: Initiating action 21: stop db_srv_stop_0 on node1 (local)
[...]
Sep 02 09:24:39 node1 pacemakerd[5994]: notice: crm_add_logfile: Additional logging available in /var/log/pacemaker.log
Sep 02 09:24:39 node1 mysql(db_srv)[6014]: ERROR: Datadir /data/mysql doesn't exist
[...]
Sep 02 09:24:39 node1 crmd[3461]: notice: te_rsc_command: Initiating action 22: start db_srv_start_0 on node1 (local)
Sep 02 09:24:39 node1 pacemakerd[6151]: notice: crm_add_logfile: Additional logging available in /var/log/pacemaker.log
Sep 02 09:24:39 node1 crmd[3461]: notice: process_lrm_event: LRM operation db_fs_monitor_10000 (call=233, rc=0, cib-update=965, confirmed=false) ok
Sep 02 09:24:39 node1 mysql(db_srv)[6210]: INFO: MySQL already running