Description of problem: I was having a volume with quota enabled and quota limit set and updated the cluster from glusterfs-3.4.0.36rhs to glusterfs-3.4.0.37rhs On a cluster on four nodes quota[1-4] on node quota1, after starting the volume, I executed the command. [root@quota1 ~]# gluster volume quota dist-rep4 list quota: Error on quota auxiliary mount (No such file or directory). this command failure we have already filed BZ 1025163 Now, I tried to execute the command on node quota3, [root@quota3 ~]# gluster volume quota dist-rep4 disable Disabling quota will delete all the quota configuration. Do you want to continue? (y/n) y quota command failed : Commit failed on 10.70.42.186. Error: Disabling quota on volume dist-rep4 has been unsuccessful ip address of quota1 ---- 10.70.42.186 whereas, actually on any node if I check the quota disabled. [root@quota2 ~]# gluster volume info dist-rep4 Volume Name: dist-rep4 Type: Distributed-Replicate Volume ID: 29b18e87-1b68-454c-894e-78294cc068d5 Status: Started Number of Bricks: 6 x 2 = 12 Transport-type: tcp Bricks: Brick1: 10.70.42.186:/rhs/brick1/d1r14 Brick2: 10.70.43.181:/rhs/brick1/d1r24 Brick3: 10.70.43.18:/rhs/brick1/d2r14 Brick4: 10.70.43.22:/rhs/brick1/d2r24 Brick5: 10.70.42.186:/rhs/brick1/d3r14 Brick6: 10.70.43.181:/rhs/brick1/d3r24 Brick7: 10.70.43.18:/rhs/brick1/d4r14 Brick8: 10.70.43.22:/rhs/brick1/d4r24 Brick9: 10.70.42.186:/rhs/brick1/d5r14 Brick10: 10.70.43.181:/rhs/brick1/d5r24 Brick11: 10.70.43.18:/rhs/brick1/d6r14 Brick12: 10.70.43.22:/rhs/brick1/d6r24 Options Reconfigured: features.quota: off features.quota-deem-statfs: on Version-Release number of selected component (if applicable): glusterfs-3.4.0.37rhs How reproducible: seen once, but first time on this build Actual results: quota disable fails Expected results: quota disable should work seamlessly Additional info: [root@quota4 ~]# gluster volume status dist-rep4 Status of volume: dist-rep4 Gluster process Port Online Pid ------------------------------------------------------------------------------ Brick 10.70.42.186:/rhs/brick1/d1r14 49192 Y 11858 Brick 10.70.43.181:/rhs/brick1/d1r24 49163 Y 18953 Brick 10.70.43.18:/rhs/brick1/d2r14 49173 Y 9832 Brick 10.70.43.22:/rhs/brick1/d2r24 49161 Y 17149 Brick 10.70.42.186:/rhs/brick1/d3r14 49193 Y 11869 Brick 10.70.43.181:/rhs/brick1/d3r24 49164 Y 18964 Brick 10.70.43.18:/rhs/brick1/d4r14 49174 Y 9843 Brick 10.70.43.22:/rhs/brick1/d4r24 49162 Y 17160 Brick 10.70.42.186:/rhs/brick1/d5r14 49194 Y 11880 Brick 10.70.43.181:/rhs/brick1/d5r24 49165 Y 18975 Brick 10.70.43.18:/rhs/brick1/d6r14 49175 Y 9854 Brick 10.70.43.22:/rhs/brick1/d6r24 49163 Y 17171 NFS Server on localhost 2049 Y 17186 Self-heal Daemon on localhost N/A Y 17190 NFS Server on 10.70.42.186 2049 Y 11894 Self-heal Daemon on 10.70.42.186 N/A Y 11898 NFS Server on 10.70.43.18 2049 Y 9868 Self-heal Daemon on 10.70.43.18 N/A Y 9873 NFS Server on 10.70.43.181 2049 Y 18987 Self-heal Daemon on 10.70.43.181 N/A Y 18994 There are no active volume tasks
From the log: [2013-10-31 14:41:48.069600] E [glusterd-utils.c:8494:glusterd_remove_auxiliary_mount] 0-management: umount on /var/run/gluster/dist-rep4/ failed, reason : Bad file descriptor [2013-10-31 14:41:48.069739] E [glusterd-op-sm.c:3693:glusterd_op_ac_commit_op] 0-management: Commit of operation 'Volume Quota' failed: -1 There is no clue on why umount failed. This happened during internal upgrade, not sure if we can re-create this problem again. Closing the bug as worksforme, please re-open if the issue is seen again.