[Gluster-users] brick errors after xfs issues

[Matthew Nicholson]

We had a brick drop out on us when xfs shutdown the filesystem due to some
errors. a prety quick check/repair/check and remounts later, the filesystem
SEEMS healthy, restarted glusterd on the system and things SEEMED okay, but
now in the brick log:


[2013-05-02 12:36:20.117170] W
[marker-quota.c:2047:mq_inspect_directory_xattr] 0-gstore-marker: cannot
add a new contribution node
[2013-05-02 12:36:20.121419] E
[marker-quota-helper.c:230:mq_dict_set_contribution]
(-->/usr/lib64/glusterfs/3.3.1/xlator/debug/io-stats.so(io_stats_lookup+0x13e)
[0x7fed88597a3e]
(-->/usr/lib64/glusterfs/3.3.1/xlator/features/marker.so(marker_lookup+0x300)
[0x7fed887at
[2013-05-02 12:36:20.122146] W
[marker-quota.c:2047:mq_inspect_directory_xattr] 0-gstore-marker: cannot
add a new contribution node
[2013-05-02 12:36:20.163452] E
[marker-quota-helper.c:230:mq_dict_set_contribution]
(-->/usr/lib64/glusterfs/3.3.1/xlator/debug/io-stats.so(io_stats_lookup+0x13e)
[0x7fed88597a3e]
(-->/usr/lib64/glusterfs/3.3.1/xlator/features/marker.so(marker_lookup+0x300)
[0x7fed887at
[2013-05-02 12:36:20.164418] W
[marker-quota.c:2047:mq_inspect_directory_xattr] 0-gstore-marker: cannot
add a new contribution node
[2013-05-02 12:36:20.165283] E
[marker-quota-helper.c:230:mq_dict_set_contribution]
(-->/usr/lib64/glusterfs/3.3.1/xlator/debug/io-stats.so(io_stats_lookup+0x13e)
[0x7fed88597a3e]
(-->/usr/lib64/glusterfs/3.3.1/xlator/features/marker.so(marker_lookup+0x300)
[0x7fed887at
[2013-05-02 12:36:20.165878] W
[marker-quota.c:2047:mq_inspect_directory_xattr] 0-gstore-marker: cannot
add a new contribution node
[2013-05-02 12:36:20.170552] E
[marker-quota-helper.c:230:mq_dict_set_contribution]
(-->/usr/lib64/glusterfs/3.3.1/xlator/debug/io-stats.so(io_stats_lookup+0x13e)
[0x7fed88597a3e]
(-->/usr/lib64/glusterfs/3.3.1/xlator/features/marker.so(marker_lookup+0x300)
[0x7fed887at
[2013-05-02 12:36:20.171489] W
[marker-quota.c:2047:mq_inspect_directory_xattr] 0-gstore-marker: cannot
add a new contribution node

endlessly.

any thoughts? I'm not finding much on this sort of error out there.

This is a 10x2 dist-replicated volume.

worst case senario, i want to pull this node (and i assume i've got to pull
its twin), whats the best way to preserve that data on there? "remove" the
nodes, and then sync the data from one brick (since they should be the
same) over to the client mount to get it on the other 9 nodes
(distributed), then wipe/re-add the 2 bricks i just removed and reblance?


--
Matthew Nicholson
Research Computing Specialist
Harvard FAS Research Computing
matthew_nicholson at harvard.edu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20130502/938806ac/attachment.html>