Opensuse 13.1 kernel patch openSUSE-2014-114 breaks ocfs2 mounting in pacemaker

Hello folks.

I applied the recent kernel update: openSUSE-2014-114 (which provides kernel-default-3.11.10-7.1.x86_64) to my 13.1 pacemaker cluster.

After rebooting and restarting the corosync, dlm and finally the pacemaker processes, my ocfs2 volume will no longer mount when tried.

2014-02-07T15:10:23.404233-05:00 wilma kernel: 1235.292697] ocfs2: Userspace control daemon is not present
2014-02-07T15:10:23.404245-05:00 wilma kernel: 1235.292705] (mount.ocfs2,5215,0):ocfs2_dlm_init:3004 ERROR: status = -3
2014-02-07T15:10:23.404246-05:00 wilma kernel: 1235.292738] (mount.ocfs2,5215,0):ocfs2_mount_volume:1881 ERROR: status = -3
2014-02-07T15:10:23.404247-05:00 wilma kernel: 1235.292776] ocfs2: Unmounting device (8,16) on (node 0)
2014-02-07T15:10:23.404247-05:00 wilma kernel: 1235.292781] (mount.ocfs2,5215,0):ocfs2_fill_super:1229 ERROR: status = -3
2014-02-07T15:10:23.404397-05:00 wilma lrmd[5016]: notice: operation_finished: fsocfs_start_0:5159:stderr mount.ocfs2: No such process while mounting /dev/sdb on /ocfs. Check ‘dmesg’ for more information on this error. ]

Suspect this latest kernel update is no longer doing things with ocfs2 as it has done with the initial 13.1 kernel - and is likely doing things as was done in 12.3.

I fell back to the previous kernel version, kernel-default-3.11.6-4.1.x86_64, and it’s working as it should.

If this is an expected result, what is the process we should be expected to use ocfs2 after applying this update?

Hello,
That is a bug , please reference this page: http://lists.opensuse.org/archive/opensuse-bugs/2013-12/msg02111.html
Good luck.

Thanks for the link. I think the resulting error is similar, but not exactly the same path to get there since that original bug references the 3.11.6 kernel and my cluster has been working fine (finally) with that version.

This issue showed up with the kernel update openSUSE-2014-114 and that update provided the kernel kernel-default-3.11.10-7.1.x86_64.

At this point I’ll work around the whole thing by not applying any more kernel updates that are pushed out by Opensuse to this cluster since it is working right now and thankfully it is not directly attached to the internet and wait for 13.2 to see if things are sorted out better by then.

On 2014-02-10 21:56, cl1420 wrote:
> and wait for 13.2 to see if things are sorted out better by
> then.

make sure you report in Bugzilla. If there is a report already, add your
info to it. If there is no report, you could wait in vain.


Cheers / Saludos,

Carlos E. R.

(from 13.1 x86_64 “Bottle” (Minas Tirith))