<div dir="ltr"><div>Hi<br></div> I saw one of our client dying after we see "gfid different on subvolume". Here is log of the client <br><div><div><br>[2014-03-03 08:30:54.286225] W
[afr-common.c:1196:afr_detect_self_heal_by_iatt] 0-bj-mams-replicate-6:
/operation/video/2014/03/03/e7/35/71/81f0a6656c077a16cad663e540543a78.pfvmeta:
gfid different on subvolume
<br>[2014-03-03 08:30:54.287017] I
[afr-self-heal-common.c:1970:afr_sh_post_nb_entrylk_gfid_sh_cbk]
0-bj-mams-replicate-6: Non blocking entrylks failed.
<br>[2014-03-03 08:30:54.287910] W [inode.c:914:inode_lookup]
(-->/usr/lib64/glusterfs/3.3.0.5rhs_iqiyi_7/xlator/debug/io-stats.so(io_stats_lookup_cbk+0xff)
[0x7fb6fd630adf]
(-->/usr/lib64/glusterfs/3.3.0.5rhs_iqiyi_7/xlator/mount/fuse.so(+0xf3f8)
[0x7fb7019da3f8]
(-->/usr/lib64/glusterfs/3.3.0.5rhs_iqiyi_7/xlator/mount/fuse.so(+0xf25b)
[0x7fb7019da25b]))) 0-fuse: inode not found<br><br>I saw similar discussion in mail list, but i don't see the solution to fix this issue.<br><a href="http://www.gluster.org/pipermail/gluster-users/2013-June/036190.html">http://www.gluster.org/pipermail/gluster-users/2013-June/036190.html</a><br>
<br></div><div>Using umount and remount, the client is alive now. but What I want know why this happen ? Is there any bugfix on this?<br>thanks <br></div></div></div>