Bug 1253309

Summary:	AFR: gluster v restart force or brick process restart doesn't heal the files
Product:	[Community] GlusterFS	Reporter:	Ravishankar N <ravishankar>
Component:	replicate	Assignee:	Ravishankar N <ravishankar>
Status:	CLOSED CURRENTRELEASE	QA Contact:
Severity:	urgent	Docs Contact:
Priority:	unspecified
Version:	mainline	CC:	annair, ashah, asrivast, bugs, gluster-bugs, ravishankar, ssampat
Target Milestone:	---	Keywords:	Reopened, Triaged
Target Release:	---
Hardware:	x86_64
OS:	Linux
Whiteboard:
Fixed In Version:	glusterfs-3.8rc2	Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:	1239021
Clones:	1255690 1256245 (view as bug list)		Environment:
Last Closed:	2016-06-16 13:31:51 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:
Bug Depends On:	1239021
Bug Blocks:	1223636, 1255690, 1256245

Description Ravishankar N 2015-08-13 12:59:54 UTC

Description of problem:

When one of the replica brick is down and do some file operation, gluster vol restart or brick process restart doesn't heal the files which needs to be healed.

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

glusterfs-3.7.1-7.el6rhs.x86_64


How reproducible:

100%

Steps to Reproduce:

1. Create 2*2 distribute replicate volume
2. Do fuse mount 
3. create some files on mount point
4. kill one of the replica brick
5. rename the file from the mount point
6. check gluster v heal <volname> info
7. restart the volume or restart the brick process


Actual results:

Files are not healed


Expected results:

volume restart or brick process restart should heal the files which need to be healed

Additional info:

Volume Name: vol0
Type: Distributed-Replicate
Volume ID: 53c64343-c537-428c-b7b7-a45f198c42a0
Status: Started
Number of Bricks: 2 x 2 = 4
Transport-type: tcp
Bricks:
Brick1: 10.70.33.214:/rhs/brick1/b001
Brick2: 10.70.33.219:/rhs/brick1/b002
Brick3: 10.70.33.225:/rhs/brick1/b003
Brick4: 10.70.44.13:/rhs/brick1/b004
Options Reconfigured:
performance.readdir-ahead: on
features.uss: enable
features.quota: on
features.inode-quota: on
features.quota-deem-statfs: on
server.allow-insecure: on
features.barrier: disable
cluster.enable-shared-storage: enable


--- Additional comment from Ravishankar N on 2015-07-03 05:45:57 EDT ---

Currently in AFR-v2, when a CHILD_UP notification is received, the index heal is triggered only on that particular child. The fix is to trigger the index heal on all local children.

While this is a bug, it is not a blocker because the files will eventually get healed in 10 minutes (default heal timeout value) or when the heal command is explicitly launched via the gluster CLI.

Comment 1 Ravishankar N 2015-08-13 13:15:30 UTC

http://review.gluster.org/#/c/11912/

Comment 2 Anand Avati 2015-08-14 09:14:59 UTC

REVIEW: http://review.gluster.org/11912 (afr: launch index heal on local subvols up on a child-up event) posted (#2) for review on master by Pranith Kumar Karampuri (pkarampu)

Comment 3 Anand Avati 2015-08-14 09:44:03 UTC

REVIEW: http://review.gluster.org/11912 (afr: launch index heal on local subvols up on a child-up event) posted (#3) for review on master by Pranith Kumar Karampuri (pkarampu)

Comment 4 Anand Avati 2015-08-21 05:29:18 UTC

REVIEW: http://review.gluster.org/11912 (afr: launch index heal on local subvols up on a child-up event) posted (#4) for review on master by Ravishankar N (ravishankar)

Comment 5 Anand Avati 2015-08-21 10:50:18 UTC

COMMIT: http://review.gluster.org/11912 committed in master by Pranith Kumar Karampuri (pkarampu) 
------
commit e4cefd6c5915dd47c6b42098236df3901665f93a
Author: Ravishankar N <ravishankar>
Date:   Thu Aug 13 18:33:08 2015 +0530

    afr: launch index heal on local subvols up on a child-up event
    
    Problem:
    When a replica's child goes down and comes up, the index heal is
    triggered only on the child that just came up. This does not serve the
    intended purpose as the list of files that need to be healed
    to this child is actually captured on the other child of the replica.
    
    Fix:
    Launch index-heal on all local children of the replica xlator which just
    received a child up. Note that afr_selfheal_childup() eventually calls
    afr_shd_index_healer() which will not run the heal on non-local
    children.
    
    Signed-off-by: Ravishankar N <ravishankar>
    
    Change-Id: Ia23e47d197f983c695ec0bcd283e74931119ee55
    BUG: 1253309
    Reviewed-on: http://review.gluster.org/11912
    Tested-by: NetBSD Build System <jenkins.org>
    Tested-by: Gluster Build System <jenkins.com>
    Reviewed-by: Krutika Dhananjay <kdhananj>
    Reviewed-by: Pranith Kumar Karampuri <pkarampu>

Comment 6 Nagaprasad Sathyanarayana 2015-10-25 15:05:12 UTC

Fix for this BZ is already present in a GlusterFS release. You can find clone of this BZ, fixed in a GlusterFS release and closed. Hence closing this mainline BZ as well.

Comment 7 Niels de Vos 2016-06-16 13:31:51 UTC

This bug is getting closed because a release has been made available that should address the reported issue. In case the problem is still not fixed with glusterfs-3.8.0, please open a new bug report.

glusterfs-3.8.0 has been announced on the Gluster mailinglists [1], packages for several distributions should become available in the near future. Keep an eye on the Gluster Users mailinglist [2] and the update infrastructure for your distribution.

[1] http://blog.gluster.org/2016/06/glusterfs-3-8-released/
[2] http://thread.gmane.org/gmane.comp.file-systems.gluster.user