Bug 1323075

Summary:	Volume status command failed during update from 3.1.2 to latest one, on non update node
Product:	Red Hat Gluster Storage	Reporter:	Byreddy <bsrirama>
Component:	glusterd	Assignee:	Atin Mukherjee <amukherj>
Status:	CLOSED NOTABUG	QA Contact:	Byreddy <bsrirama>
Severity:	high	Docs Contact:
Priority:	unspecified
Version:	rhgs-3.1	CC:	rhs-bugs, sasundar, storage-qa-internal, vbellur
Target Milestone:	---	Keywords:	ZStream
Target Release:	---
Hardware:	x86_64
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2016-04-01 09:08:19 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:

Description Byreddy 2016-04-01 08:10:05 UTC

Description of problem:
=======================
Volume status command  failed on not updated node with error message "Commit failed on <host name of updated node>


Version-Release number of selected component (if applicable):
=============================================================
glusterfs-3.7.9-1.el7


How reproducible:
==================
One time


Steps to Reproduce:
===================
1.Have two node cluster with rhgs 3.1.2 (node-1 and node-2)
2.Create and start  2*2 volume and do FUSE mount and start IO.
3.Update one of the node (say node-1) to latest 3.1.3 //follow in-service update procedure
4.Check volume status on node-2 multiple times

Actual results:
===============
Volume status on node-2 failed one time with error message "Commit failed on <host name of node-1>.


Expected results:
=================
Volume status command should not fail.


Additional info:
==================

Useful log info can be found below to debug

Comment 3 Atin Mukherjee 2016-04-01 09:08:19 UTC

First of all you'd need to provide log for dhcp37-171.lab.eng.blr.redhat.com to understand why the commit failed and more over this is not a supported configuration where all the nodes are not upgraded as we don't support hybrid cluster.

If required you can test this upstream and if you see the same issue file a bug upstream since we support heterogeneous cluster there, but in downstream this is not valid.