+++ This bug was initially created as a clone of Bug #1297695 +++ Description of problem: https://www.gluster.org/pipermail/gluster-users.old/2015-December/024568.html : Here's a snippet of the user's feedback on the thread: <snip> ... ... ... Also, it would be great if the heal info command could return faster, sometimes it takes over a minute. ... ... </snip> Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info:
Patch merged on master: http://review.gluster.org/#/c/13326/ Will keep the bug in 'POST' state.
QATP for the fix and the result status: ====================================== TC#1:heal info command should throw the output immediatly(within a second) for a newly created afr volume 1. create a 6x2 volume and start it 2. Now issue a "gluster v heal info" cmnd on this volume. Expected Result: the o/p must not hang and must throw the o/p in less than a second Result-->PASS(approx time taken by heal info to display and exit was 0.4 Sec) TC#2:heal info should not hang on to a file where IO is happening, but must rather throw the o/p and exit 1. create a 6x2 volume and start it and mount it 2. Now create a file using dd and keep writing to it, do continuous IO to it 3. While IOs are happening issue heal info command EXPECTED result: the heal info must throw the o/p and exit without hanging as long as the IO is happening Result-->PASS(approx time taken by heal info to display and exit was 0.5 Sec) TC#3:heal info should not hang on to a file where IO is happening and one of the brick of the file getting hashed to is brought down, but must rather throw the o/p and exit 1. create a 6x2 volume and start it and mount it 2. Now create a file using dd and keep writing to it, do continuous IO to it 3. While IOs are happening, bring down the one of the brick where the file is getting hashed to 4. Now wait for say about 5 min while the IOs while IO keeps happening and issue heal info command EXPECTED result: the heal info must throw the o/p and exit without hanging as long as the IO is happening Result-->PASS (approx time taken by heal info to display and exit was 0.6Sec) TC#4:heal info should not hang while there are lot of files to be healed 1. create a 6x2 volume and start it and mount it 2. Now start a linux untar on the mount 3. while untar is happening, bring down about 3 bricks and wait for about 3 min 4. Now issue a heal info on the volume Expected Result: the heal output should be displaying in running style for all the files needed to heal and must not get stuck or hang during anytime Result-->PASS, however, the o/p command used to get stuck in between for some files for about 15 seconds, discussed with dev and told that could be expected as long as TC#1 ,TC#2 and TC#3 pass version of validation: ====================== 3.7.9-2
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/RHBA-2016:1240