Bug 1542380

Summary:	Changes to self-heal logic w.r.t. detecting of split-brains
Product:	[Community] GlusterFS	Reporter:	Ravishankar N <ravishankar>
Component:	replicate	Assignee:	Ravishankar N <ravishankar>
Status:	CLOSED CURRENTRELEASE	QA Contact:
Severity:	unspecified	Docs Contact:
Priority:	unspecified
Version:	4.0	CC:	bugs, olim
Target Milestone:	---	Keywords:	Triaged
Target Release:	---
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:	glusterfs-4.0.0	Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:	1539358	Environment:
Last Closed:	2018-03-15 11:26:23 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:	1539358, 1597123
Bug Blocks:	1384983, 1541458

Description Ravishankar N 2018-02-06 08:56:55 UTC

+++ This bug was initially created as a clone of Bug #1539358 +++

Description of problem:
    We currently don't have a roll-back/undoing of post-ops if quorum is not
    met. Though the FOP is still unwound with failure, the xattrs remain on
    the disk.  Due to these partial post-ops and partial heals (healing only when
    2 bricks are up), we can end up in split-brain purely from the afr
    xattrs point of view i.e each brick is blamed by atleast one of the
    others. These scenarios are hit when there is frequent
    connect/disconnect of the client/shd to the bricks.

--- Additional comment from Worker Ant on 2018-01-28 03:40:54 EST ---

REVIEW: https://review.gluster.org/19349 (afr: don't treat false cases of split-brain as genuine) posted (#1) for review on master by Ravishankar N

--- Additional comment from Worker Ant on 2018-02-01 09:18:21 EST ---

COMMIT: https://review.gluster.org/19349 committed in master by "Ravishankar N" <ravishankar> with a commit message- afr: don't treat all cases all bricks being blamed as split-brain

Problem:
We currently don't have a roll-back/undoing of post-ops if quorum is not
met. Though the FOP is still unwound with failure, the xattrs remain on
the disk.  Due to these partial post-ops and partial heals (healing only when
2 bricks are up), we can end up in split-brain purely from the afr
xattrs point of view i.e each brick is blamed by atleast one of the
others. These scenarios are hit when there is frequent
connect/disconnect of the client/shd to the bricks while I/O or heal
are in progress.

Fix:
Instead of undoing the post-op, pick a source based on the xattr values.
If 2 bricks blame one, the blamed one must be treated as sink.
If there is no majority, all are sources. Once we pick a source,
self-heal will then do the heal instead of erroring out due to
split-brain.

Change-Id: I3d0224b883eb0945785ade0e9697a1c828aec0ae
BUG: 1539358
Signed-off-by: Ravishankar N <ravishankar>

Comment 1 Worker Ant 2018-02-06 08:57:53 UTC

REVIEW: https://review.gluster.org/19505 (afr: don't treat all cases all bricks being blamed as split-brain) posted (#1) for review on release-4.0 by Ravishankar N

Comment 2 Worker Ant 2018-02-06 14:25:43 UTC

COMMIT: https://review.gluster.org/19505 committed in release-4.0 by "Shyamsundar Ranganathan" <srangana> with a commit message- afr: don't treat all cases all bricks being blamed as split-brain

Problem:
We currently don't have a roll-back/undoing of post-ops if quorum is not
met. Though the FOP is still unwound with failure, the xattrs remain on
the disk.  Due to these partial post-ops and partial heals (healing only when
2 bricks are up), we can end up in split-brain purely from the afr
xattrs point of view i.e each brick is blamed by atleast one of the
others. These scenarios are hit when there is frequent
connect/disconnect of the client/shd to the bricks while I/O or heal
are in progress.

Fix:
Instead of undoing the post-op, pick a source based on the xattr values.
If 2 bricks blame one, the blamed one must be treated as sink.
If there is no majority, all are sources. Once we pick a source,
self-heal will then do the heal instead of erroring out due to
split-brain.

Change-Id: I3d0224b883eb0945785ade0e9697a1c828aec0ae
BUG: 1542380
Signed-off-by: Ravishankar N <ravishankar>
(cherry picked from commit 0e6e8216823c2d9dafb81aae0f6ee3497c23d140)

Comment 3 Shyamsundar 2018-03-15 11:26:23 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-4.0.0, please open a new bug report.

glusterfs-4.0.0 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://lists.gluster.org/pipermail/announce/2018-March/000092.html
[2] https://www.gluster.org/pipermail/gluster-users/