downstream patch is merged now.
*** Bug 1331164 has been marked as a duplicate of this bug. ***
Update: ======== Build Used: glusterfs-server-3.12.2-7.el7rhgs.x86_64 Verified below scenario's Scenario 1: 1. Create 1x2 volume and disable shd. 2. Create a FILE and chown it to user-1 3. Kill one brick. 4. As user-1 (non root), do chmod on FILE so that there is pending metadata heal. 5. Bring the down brick up. 6. As another user-2, access the same FILE 7. healing should be successful Scenario 2: 1. Create 2 * 3 volume and disable shd 2. create 500 files 3. chown for 1st 250 files to "qa_func" 4. chown for last 250 files to "qa_perf" 5. kill one of the brick 6. As "qa_func" user, change permissions for the 1st 250 files 7. As "qa_perf" user, change permissions for the last 250 files 8. Bring up the down brick 9. Do a lookup from 3rd user ( "qa_all" ) on all the files 10. Healing should be successful In Both Scenarios, healing is success after lookup happened ( from another user ) Changing status to Verified
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://access.redhat.com/errata/RHSA-2018:2607