[Linux-cluster] panic after "jid=0, already locked for use"
rpeterso at redhat.com
Thu Dec 11 15:05:08 UTC 2008
----- "Nathan Stratton" <nathan at robotics.net> wrote:
| Is there anyway to save my data? fsck runs clean, but I am still
| the jid=0, already locked for use and then the panic.
| Dec 11 07:56:28 xen0 kernel: GFS2: fsid=: Trying to join cluster
| "lock_nolock", "xen_sjc:share"
| Dec 11 07:56:28 xen0 kernel: GFS2: fsid=xen_sjc:share.0: Joined
| Now mounting FS...
| Dec 11 07:56:28 xen0 kernel: GFS2: fsid=xen_sjc:share.0: jid=0,
| locked for use
| Dec 11 07:56:28 xen0 kernel: GFS2: fsid=xen_sjc:share.0: jid=0:
| Looking at
| Dec 11 07:56:29 xen0 kernel: GFS2: fsid=xen_sjc:share.0: fatal:
| consistency error
| Dec 11 07:56:29 xen0 kernel: GFS2: fsid=xen_sjc:share.0: inode = 4
| Dec 11 07:56:29 xen0 kernel: GFS2: fsid=xen_sjc:share.0: function =
| jhead_scan, file = fs/gfs2/recovery.c, line = 239
I think you may have hit bugzilla bug 457557. I'm not sure if you
can view the record, due to permissions (which is out of my
control), but here is a link to try:
In the case of the bugzilla, I helped the guy who reported it patch
his file system with the gfs2_edit and everything was okay again.
It was not difficult. Perhaps what you should do is:
gfs2_edit -p journal0 /dev/your/device > /tmp/j0
Then use some tool like "less" to search /tmp/j0 for "---" which
will indicate where that journal wrapped. Then examine the
sequence numbers printed out and see if you can find a duplicate.
In the bugzilla case, I had the reporter patch the sequence
numbers so there were no duplicates, and gfs2 was able to
mount happily again. I'm willing to help with the process if
need be. If this is the same problem, I never did figure out
how the duplicate sequence number got in there, but you're only
the second person to have encountered that problem.
I will eventually get to the bottom of it though; it's a matter
of juggling priorities.
Also, I'm not sure what version of gfs2 you're running, but GFS2
was not previously considered ready for production use yet,
although it hasn't changed much recently. Nevertheless, you may
want to make sure you've got the latest and greatest GFS2 kernel
code (from source.) If it's the problem listed in the bz, newer
code would not have saved you from hitting the problem though.
Red Hat GFS
More information about the Linux-cluster