[Bugs] [Bug 1330450] New: [geo-rep]: schedule_georep.py doesn't touch the mount in every iteration

[bugzilla at redhat.com]

https://bugzilla.redhat.com/show_bug.cgi?id=1330450

            Bug ID: 1330450
           Summary: [geo-rep]: schedule_georep.py doesn't touch the mount
                    in every iteration
           Product: GlusterFS
           Version: 3.7.11
         Component: geo-replication
          Severity: urgent
          Assignee: bugs at gluster.org
          Reporter: avishwan at redhat.com
                CC: avishwan at redhat.com, bugs at gluster.org,
                    chrisw at redhat.com, csaba at redhat.com,
                    nlevinki at redhat.com, rhinduja at redhat.com,
                    storage-qa-internal at redhat.com
        Depends On: 1328397, 1328399



+++ This bug was initially created as a clone of Bug #1328399 +++

+++ This bug was initially created as a clone of Bug #1328397 +++

Description of problem:
=======================

Ran the script while there was no IO inprogress, checkpoint never reached for
few of the active workers and eventually the script never completed. The reason
is not to touch the mount point in every iteration. 

Modified script provided by dev works:

[root at dhcp37-182 ~]# diff /usr/share/glusterfs/scripts/schedule_georep.py
/tmp/schedule_georep.py
134d133
<              "--xlator-option=\"*dht.lookup-unhashed=off\"",
138d136
<              "--client-pid=-1",
142d139
< 
148c145
<     #cleanup(hostname, volname, mnt)
---
>     cleanup(hostname, volname, mnt)
416,422d412
<             if not summary["checkpoints_ok"]:
<                 # If Checkpoint is not complete after a iteration means brick
<                 # was down and came online now. SETATTR on mount is not
<                 # recorded, So again issue touch on mount root So that
<                 # Stime will increase and Checkpoint will complete.
<                 touch_mount_root(args.mastervol)
< 
432a423,428
>         else:
>             # If Checkpoint is not complete after a iteration means brick
>             # was down and came online now. SETATTR on mount is not
>             # recorded, So again issue touch on mount root So that
>             # Stime will increase and Checkpoint will complete.
>             touch_mount_root(args.mastervol)
[root at dhcp37-182 ~]# 

Version-Release number of selected component (if applicable):
==============================================================

glusterfs-3.7.9-1.el7rhgs.x86_64

How reproducible:
=================

1/1

Steps to Reproduce:
===================
1. Create data on master volume (6x2)
2. Create geo-rep session
3. Run the script

--- Additional comment from Vijay Bellur on 2016-04-19 06:33:45 EDT ---

REVIEW: http://review.gluster.org/14029 (geo-rep: Fix checkpoint issue in
scheduler) posted (#1) for review on master by Aravinda VK
(avishwan at redhat.com)

--- Additional comment from Vijay Bellur on 2016-04-20 06:59:08 EDT ---

REVIEW: http://review.gluster.org/14029 (geo-rep: Fix checkpoint issue in
scheduler) posted (#2) for review on master by Aravinda VK
(avishwan at redhat.com)

--- Additional comment from Vijay Bellur on 2016-04-22 03:05:27 EDT ---

REVIEW: http://review.gluster.org/14029 (geo-rep: Fix checkpoint issue in
scheduler) posted (#3) for review on master by Aravinda VK
(avishwan at redhat.com)

--- Additional comment from Vijay Bellur on 2016-04-26 05:14:32 EDT ---

COMMIT: http://review.gluster.org/14029 committed in master by Aravinda VK
(avishwan at redhat.com) 
------
commit 8590c1cf3c27468177c425c920cab01f52b251e5
Author: Aravinda VK <avishwan at redhat.com>
Date:   Tue Apr 19 15:30:19 2016 +0530

    geo-rep: Fix checkpoint issue in scheduler

    If checkpoint is not met, Scheduler script should touch the
    Mount point so that SETATTR will get recorded in every brick
    Changelog. Script was not touching the mount point in each iteration.

    BUG: 1328399
    Change-Id: I2718a764fb3e550742c9dcd316724683561ddf18
    Signed-off-by: Aravinda VK <avishwan at redhat.com>
    Reviewed-on: http://review.gluster.org/14029
    Smoke: Gluster Build System <jenkins at build.gluster.com>
    Reviewed-by: Kotresh HR <khiremat at redhat.com>
    CentOS-regression: Gluster Build System <jenkins at build.gluster.com>
    NetBSD-regression: NetBSD Build System <jenkins at build.gluster.org>


Referenced Bugs:

https://bugzilla.redhat.com/show_bug.cgi?id=1328397
[Bug 1328397] [geo-rep]: schedule_georep.py doesn't touch the mount in
every iteration
https://bugzilla.redhat.com/show_bug.cgi?id=1328399
[Bug 1328399] [geo-rep]: schedule_georep.py doesn't touch the mount in
every iteration
-- 
You are receiving this mail because:
You are on the CC list for the bug.
You are the assignee for the bug.