Upstream patch: http://review.gluster.org/#/c/12386/

upstream patch:    http://review.gluster.org/#/c/12386/
release-3.7 patch: http://review.gluster.org/#/c/12487/
downstream patch: https://code.engineering.redhat.com/gerrit/60609

Verification failed.
Able to reproduce this bug again on build glusterfs-libs-3.7.5-6.el7rhgs.x86_64

[root@localhost newtest]# gluster v quota testvol disable
Disabling quota will delete all the quota configuration. Do you want to continue? (y/n) y
^V^C

[root@localhost newtest]# gluster v info
 
Volume Name: testvol
Type: Distributed-Replicate
Volume ID: 6e5023e1-5942-4146-933e-3d06d6921029
Status: Started
Number of Bricks: 2 x 2 = 4
Transport-type: tcp
Bricks:
Brick1: 10.70.47.143:/rhs/brick3/b1
Brick2: 10.70.47.145:/rhs/brick3/b2
Brick3: 10.70.47.2:/rhs/brick3/b3
Brick4: 10.70.47.3:/rhs/brick3/b4
Options Reconfigured:
features.inode-quota: off
features.quota: off
performance.readdir-ahead: on

[root@localhost newtest]# gluster v quota testvol enable

[root@localhost newtest]# gluster v quota testvol limit-usage / 2GB
quota command failed : Failed to find the directory /var/run/gluster/testvol/. Reason : No such file or directory

Hi Anil,

You have mentioned two problems in comment# 5

Problem 1)
[root@localhost newtest]# gluster v quota testvol disable
Disabling quota will delete all the quota configuration. Do you want to continue? (y/n) y
^V^C

[root@localhost newtest]# gluster v info
Options Reconfigured:
features.inode-quota: off
features.quota: off

Once you execute command 'gluster v quota testvol disable', CLI will send the request to glusterd to perform the operation and waits for the result response from glusterd. 
If you kill CLI it will not receive and response, but glusterd will continue to execute the operation you initiated.


Problem 2)
[root@localhost newtest]# gluster v quota testvol enable

[root@localhost newtest]# gluster v quota testvol limit-usage / 2GB
quota command failed : Failed to find the directory /var/run/gluster/testvol/. Reason : No such file or directory

Need to investigate wht limit-usage on '/' is failing. File a new new bug for this 2nd problem I will investigate.

Consider below example
1) When a quota is disabled, glusterd will initiate a backend process 'find / | xargs setxattr -x ....'
   to clean up all quota xattrs in the backend.
2) If there is a file
   /dir/a (if a is 1MB, file a and /dir and / will be accounted 1MB)
3) If a clanup process clean xattr from file a, but failed to cleanup xattr on /dir and / (may be node rebooted or process terminated)
4) now if quota is enabled
   file 'a' will be accounted 1MB, this 1MB is added to /dir and /
   /dir and / already has 1MB accounted which was not cleaned and after adding 1MB the quota usage becomes 2MB which is wrong...

To prevent this problem version number is added to the quota xattrs (suffixed in key)and this number is incremented on every enable operation

On first enable:
trusted.glusterfs.quota.size.1=0x0000000008c00000000000000000000e0000000000000001
trusted.glusterfs.quota.limit-set.1=0x0000000100000000ffffffffffffffff
trusted.glusterfs.quota.contri.0x0000000000000001.1=0x0000000008c00000000000000000000e0000000000000001

On second enable:
trusted.glusterfs.quota.size.2=0x0000000008c00000000000000000000e0000000000000001
trusted.glusterfs.quota.limit-set.2=0x0000000100000000ffffffffffffffff
trusted.glusterfs.quota.contri.0x0000000000000001.2=0x0000000008c00000000000000000000e0000000000000001

As discussed with Development team, Disabling and enabling will increment the trusted.glusterfs.quota.limit-set.2.

Marking this bug verified on build glusterfs-3.7.5-6.el7rhgs.x86_64

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://rhn.redhat.com/errata/RHBA-2016-0193.html