Why did my clustered service take a long time to fail or recover when the resources have the "__independent_subtree" attribute enabled on RHEL 5 or RHEL 6?

Issue

• Why did my clustered service take a long time to fail when the resources have the __independent_subtree attribute enabled?
• When my service failed, I see it performing an inline restart of resources over and over, many more times than the __max_restarts attribute is set to

Oct 31 10:02:19 rgmanager [fs] Checking fs "/mnt/ext3vol1", Level 20
Oct 31 10:02:19 rgmanager [fs] fs:/mnt/ext3vol1: /dev/mapper/vgTarget2-lvol1 is not mounted on /mnt/ext3vol1
Oct 31 10:02:20 rgmanager Some independent resources in service:demo failed; Attempting inline recovery
Oct 31 10:02:30 rgmanager [fs] 'mount -t ext3  /dev/dm-2 /mnt/ext3vol1' failed, error=32
Oct 31 10:02:33 rgmanager start on fs "/mnt/ext3vol1" returned 1 (generic error)
Oct 31 10:02:33 rgmanager Inline recovery of service:demo complete
Oct 31 10:02:40 rgmanager [fs] Checking fs "/mnt/ext3vol1", Level 10
Oct 31 10:02:41 rgmanager [fs] fs:/mnt/ext3vol1: /dev/mapper/vgTarget2-lvol1 is not mounted on /mnt/ext3vol1
Oct 31 10:02:41 rgmanager Some independent resources in service:demo failed; Attempting inline recovery
Oct 31 10:02:49 rgmanager [fs] 'mount -t ext3  /dev/dm-2 /mnt/ext3vol1' failed, error=32
Oct 31 10:02:52 rgmanager start on fs "/mnt/ext3vol1" returned 1 (generic error)
Oct 31 10:02:52 rgmanager Inline recovery of service:demo complete
Oct 31 10:02:59 rgmanager [ip] Checking 192.168.122.221, Level 10
[... over and over ...]