This bug has been copied from bug #555901 and has been proposed to be backported to 4.8 z-stream (EUS).
http://git.fedorahosted.org/git/?p=cluster.git;a=commit;h=04d4b4acc681f8af1a467ffac077e5898ad19592
This patch cannot kill processes running directly on the mountpoint. For more info please see the parent bug 555901.
http://git.fedorahosted.org/git?p=cluster.git;a=commit;h=26a1b3ec5c6831fda8e46f5e0ff133c5605f91eb
The patch has still problems handling SIGTERM ignoring applications on clusterfs with force_umount="1". <rm> <resources> <clusterfs device="/dev/vedder/vedder0" force_unmount="1" self_fence="0" fstype="gfs" mountpoint="/mnt/vedder0" name="vedderfs" options=""/> </resources> <service autostart="1" name="jkservice"> <clusterfs ref="vedderfs"/> </service> </rm> Run simple script from mountpoint: #!/bin/bash trap "" SIGTERM; sleep 10000; Relocating the service will fail: Jun 28 12:05:28 z2 clurgmgrd[12176]: <notice> Stopping service jkservice Jun 28 12:05:32 z2 clurgmgrd: [12176]: <err> 'umount /mnt/vedder0' failed, error=0 Jun 28 12:05:32 z2 clurgmgrd[12176]: <notice> stop on clusterfs "vedderfs" returned 2 (invalid argument(s)) Jun 28 12:05:32 z2 clurgmgrd[12176]: <crit> #12: service:jkservice failed to stop; intervention required Jun 28 12:05:32 z2 clurgmgrd[12176]: <notice> Service jkservice is failed
http://git.fedorahosted.org/git/?p=cluster.git;a=commit;h=26a1b3ec5c6831fda8e46f5e0ff133c5605f91eb Replaced previous patch with the above patch.
Tested on clusterfs and fs. All processess accessing the mountpoints were killed and none of the others (including those running in the example given by reporter) were killed.
Technical note added. If any revisions are required, please edit the "Technical Notes" field accordingly. All revisions will be proofread by the Engineering Content Services team. New Contents: If an application was using a mount point similar to a mount point managed by rgmanager and force_unmount was used, the file system agent could kill that particular process. This issue has been resolved and the file system does not kill the process anymore.
An advisory has been issued which should help the problem described in this bug report. This report is therefore being closed with a resolution of ERRATA. For more information on therefore solution and/or where to find the updated files, please follow the link below. You may reopen this bug report if the solution does not work for you. http://rhn.redhat.com/errata/RHBA-2010-0550.html