Bug 995811

Summary:	corosync memory leaks when the member nodes stop/start/restart
Product:	Red Hat Enterprise Linux 6	Reporter:	youyun <youyunyehe>
Component:	corosync	Assignee:	Jan Friesse <jfriesse>
Status:	CLOSED INSUFFICIENT_DATA	QA Contact:	Cluster QE <mspqa-list>
Severity:	high	Docs Contact:
Priority:	unspecified
Version:	6.0	CC:	ccaulfie, cluster-maint, fdinitto, rjoost, youyunyehe
Target Milestone:	rc
Target Release:	---
Hardware:	ia64
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2014-02-13 10:53:22 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:

Description youyun 2013-08-11 09:53:30 UTC

Description of problem:
the corosync memory leaks when the peer node stop/start/restart

Version-Release number of selected component (if applicable):
corosync-1.4.6
REHL6.0

How reproducible:
create one cluster, let one or more members stop/start/restart
then `ps auxf | grep corosync` you will find that the memory is increaments 
every time the other nodes stop/start/restart 

e.g
NODEA
while [ 1 ];do /etc/init.d/corosync restart;done; 

NODEB
ps auxf | grep corosync 

in NODEB, you will find the memory increaments very fast and will not down forever.

Steps to Reproduce:
1. make two nodes cluster
2. let one node stop/start/restart
3. `ps -auxf | grep corosync` to see the node which not stop/start/restart 


Actual results:
the memory which consumed by corosync increaments very fast!

Expected results:
the memory will not leak

Additional info:

Comment 2 Jan Friesse 2013-08-12 06:35:10 UTC

Youyun,
can you please attach corosync.conf? Are there any services using corosync running or it's pure corosync start/stop test? If so, are you sure corosync will really never stop eating memory (so every restart means new allocated memory)? Because if it's not so, it's not a bug (there are structures, which may be filled during lifetime of corosync but they are limited in size).

Also 6.5 capacity is constrained so moving to 6.6.

Comment 3 Jan Friesse 2013-11-13 08:31:14 UTC

Youyun,
can you please attach additional files/provide additional informations I've asked in Aug? Otherwise I will probably close this bug, because I'm unable to reproduce it.

Thanks,
  Honza

Comment 4 Jan Friesse 2014-01-06 13:30:07 UTC

Youyun,
any news?

Comment 5 Red Hat Bugzilla 2023-09-14 01:49:02 UTC

The needinfo request[s] on this closed bug have been removed as they have been unresolved for 1000 days