Bug 1489511

Summary:	return ENOSYS for 'non readable' FOPs
Product:	[Community] GlusterFS	Reporter:	Ravishankar N <ravishankar>
Component:	arbiter	Assignee:	Ravishankar N <ravishankar>
Status:	CLOSED CURRENTRELEASE	QA Contact:
Severity:	unspecified	Docs Contact:
Priority:	unspecified
Version:	3.12	CC:	bugs, ndevos
Target Milestone:	---	Keywords:	Triaged
Target Release:	---
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:	glusterfs-glusterfs-3.12.1	Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:	1484722	Environment:
Last Closed:	2017-09-14 07:43:25 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:	1484722
Bug Blocks:	1488560

Description Ravishankar N 2017-09-07 15:03:55 UTC

+++ This bug was initially created as a clone of Bug #1484722 +++

Description of problem:
feedback at https://review.gluster.org/#/c/18103/1//COMMIT_MSG@15 :

AFR marks the arbiter brick as 'non readable', so any read-based FOP won't be sent on it.(See: http://review.gluster.org/13539). So I don't think this change is really needed. After 13539, even arbiter_readv() won't be hit as well.

It might be a good idea to have this defensive fix (unwind the FOP with failure) in the rogue case AFR does wind the read fop to arbiter.


Expected results:
No change in behaviour, AFR should not call the FOPs from the arbiter xlator in any case.

--- Additional comment from Worker Ant on 2017-08-24 04:28:42 EDT ---

REVIEW: https://review.gluster.org/18103 (arbiter: return ENOTCONN for 'non readable' FOPs) posted (#2) for review on master by Niels de Vos (ndevos)

--- Additional comment from Worker Ant on 2017-09-04 09:08:48 EDT ---

REVIEW: https://review.gluster.org/18103 (arbiter: return ENOTCONN for 'non readable' FOPs) posted (#3) for review on master by Ravishankar N (ravishankar)

--- Additional comment from Worker Ant on 2017-09-05 02:00:22 EDT ---

REVIEW: https://review.gluster.org/18103 (arbiter: return ENOTCONN for 'non readable' FOPs) posted (#4) for review on master by Ravishankar N (ravishankar)

--- Additional comment from Worker Ant on 2017-09-05 04:36:47 EDT ---

REVIEW: https://review.gluster.org/18103 (arbiter: return ENOSYS for 'non readable' FOPs) posted (#5) for review on master by Niels de Vos (ndevos)

--- Additional comment from Worker Ant on 2017-09-05 10:03:43 EDT ---

COMMIT: https://review.gluster.org/18103 committed in master by Ravishankar N (ravishankar) 
------
commit b1352d0974328b367afa7360e9523585efb7178d
Author: Niels de Vos <ndevos>
Date:   Wed Aug 23 18:18:24 2017 +0200

    arbiter: return ENOSYS for 'non readable' FOPs
    
    AFR marks the arbiter as 'non readable'. This has been introduced with
    commit 8ab87137 (afr: do not set arbiter as a readable subvol in inode
    context). arbiter_readv() should not get called anymore, so it could be
    removed. However, it is a good defensive approach to have all the inode
    read FOPs that can not be handled by the arbiter to return ENOSYS.
    
    Change-Id: I6ea41680832859bd6790dc8d7440ee98d38205fc
    BUG: 1484722
    Signed-off-by: Niels de Vos <ndevos>
    Signed-off-by: Ravishankar N <ravishankar>
    Reviewed-on: https://review.gluster.org/18103
    Smoke: Gluster Build System <jenkins.org>
    CentOS-regression: Gluster Build System <jenkins.org>

Comment 1 Worker Ant 2017-09-07 15:04:36 UTC

REVIEW: https://review.gluster.org/18227 (arbiter: return ENOSYS for 'non readable' FOPs) posted (#1) for review on release-3.12 by Ravishankar N (ravishankar)

Comment 2 Worker Ant 2017-09-10 18:34:42 UTC

COMMIT: https://review.gluster.org/18227 committed in release-3.12 by jiffin tony Thottan (jthottan) 
------
commit 143714d96eff50501b1a5a3debf794cae9f91005
Author: Niels de Vos <ndevos>
Date:   Wed Aug 23 18:18:24 2017 +0200

    arbiter: return ENOSYS for 'non readable' FOPs
    
    AFR marks the arbiter as 'non readable'. This has been introduced with
    commit 8ab87137 (afr: do not set arbiter as a readable subvol in inode
    context). arbiter_readv() should not get called anymore, so it could be
    removed. However, it is a good defensive approach to have all the inode
    read FOPs that can not be handled by the arbiter to return ENOSYS.
    
    > Reviewed-on: https://review.gluster.org/18103
    > Smoke: Gluster Build System <jenkins.org>
    > CentOS-regression: Gluster Build System <jenkins.org>
    (cherry picked from commit b1352d0974328b367afa7360e9523585efb7178d)
    
    Change-Id: I6ea41680832859bd6790dc8d7440ee98d38205fc
    BUG: 1489511
    Signed-off-by: Niels de Vos <ndevos>
    Signed-off-by: Ravishankar N <ravishankar>
    Reviewed-on: https://review.gluster.org/18227
    Smoke: Gluster Build System <jenkins.org>
    CentOS-regression: Gluster Build System <jenkins.org>
    Reviewed-by: Pranith Kumar Karampuri <pkarampu>

Comment 3 Jiffin 2017-09-14 07:43:25 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-glusterfs-3.12.1, please open a new bug report.

glusterfs-glusterfs-3.12.1 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/gluster-users/2017-September/032441.html
[2] https://www.gluster.org/pipermail/gluster-users/