Bug 1037274

Summary:	network disconnect/reconnect does not resume data access to server
Product:	Red Hat Gluster Storage	Reporter:	Anand Avati <aavati>
Component:	glusterfs	Assignee:	Vijaikumar Mallikarjuna <vmallika>
Status:	CLOSED ERRATA	QA Contact:	Sachidananda Urs <surs>
Severity:	high	Docs Contact:
Priority:	high
Version:	unspecified	CC:	chrisw, gluster-bugs, grajaiya, psriniva, smohan, vagarwal, vbellur, vmallika
Target Milestone:	---	Keywords:	ZStream
Target Release:	RHGS 2.1.2
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:	glusterfs-3.4.0.49rhs-1.el6rhs	Doc Type:	Bug Fix
Doc Text:	Previously, on a FUSE mount point the size of the volume was displayed incorrectly when any brick went off-line and came back online subsequently. With this fix, the network reconnection issue is fixed.	Story Points:	---
Clone Of:	1037267	Environment:
Last Closed:	2014-02-25 08:07:12 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:
Bug Depends On:	1037267
Bug Blocks:

Description Anand Avati 2013-12-03 03:50:42 UTC

+++ This bug was initially created as a clone of Bug #1037267 +++

Description of problem:

When the network to a server goes down (pull out network cable, or iptables -j DROP) and comes back up, if the rpc client had witnessed a disconnect then the reconnection to the server is always incomplete

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


How reproducible:

always


Steps to Reproduce:
0. Create a distribute volume, mount a client, run df and note size
1. On one server, run iptables -I INPUT -p tcp --dport 111:65535 -j DROP
2. Run df on client and wait for size to decrease
3. iptables -F to "resume" network,

Actual results:

df never shows original size even after waiting however long and retrying many times

Expected results:

df should show original size indicating connectivity to the server which was down

Additional info:

Logs have the line "Server and Client lk-version numbers are same, no need to reopen the fds"

--- Additional comment from Anand Avati on 2013-12-02 22:38:30 EST ---

REVIEW: http://review.gluster.org/6396 (protocol/client: handle network disconnect/reconnect properly) posted (#1) for review on master by Anand Avati (avati)

Comment 1 Sachidananda Urs 2013-12-18 10:57:29 UTC

Verified as per Comment 0. This works as expected. Used and available space on df are reflected as expected when iptables rules are flushed.

Comment 2 Pavithra 2014-01-03 06:53:07 UTC

Please verify the edited doc text for technical accuracy.

Comment 3 Vijaikumar Mallikarjuna 2014-01-03 07:23:30 UTC

doctext looks good to me.

Comment 5 errata-xmlrpc 2014-02-25 08:07:12 UTC

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

http://rhn.redhat.com/errata/RHEA-2014-0208.html