Bug 755509

Summary:	possible memory leak
Product:	[Fedora] Fedora	Reporter:	Oliver Henshaw <oliver.henshaw>
Component:	irqbalance	Assignee:	Anton Arapov <anton>
Status:	CLOSED CURRENTRELEASE	QA Contact:	Fedora Extras Quality Assurance <extras-qa>
Severity:	unspecified	Docs Contact:
Priority:	unspecified
Version:	16	CC:	anton, jonabbey, nobody, trevor
Target Milestone:	---
Target Release:	---
Hardware:	x86_64
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2011-12-12 11:03:10 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:

Description Oliver Henshaw 2011-11-21 10:57:53 UTC

Description of problem:

I have been running the system for a couple of days (not sure how long it's actually been awake) and noticed irqbalance is using 17M in top.

This looks like http://code.google.com/p/irqbalance/issues/detail?id=22 - fixed upstream.


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

irqbalance-1.0.2-1.fc16.x86_64


Additional info:

# ps aux | grep irqbalance
root      1158  0.0  0.4  38116 15464 ?        Ss   Nov18   0:08 /usr/sbin/irqbalance
# pmap `pidof irqbalance`
1158:   /usr/sbin/irqbalance
0000000000400000     28K r-x--  /usr/sbin/irqbalance (deleted)
0000000000606000      4K r----  /usr/sbin/irqbalance (deleted)
0000000000607000      4K rw---  /usr/sbin/irqbalance (deleted)
0000000002111000    240K rw---    [ anon ]
000000000214d000  17160K rw---    [ anon ]
0000003ba7800000    136K r-x--  /lib64/ld-2.14.90.so
0000003ba7a21000      4K r----  /lib64/ld-2.14.90.so
0000003ba7a22000      4K rw---  /lib64/ld-2.14.90.so
0000003ba7a23000      4K rw---    [ anon ]
0000003ba7c00000   1144K r-x--  /lib64/libglib-2.0.so.0.3000.1
0000003ba7d1e000   2044K -----  /lib64/libglib-2.0.so.0.3000.1
0000003ba7f1d000      4K r----  /lib64/libglib-2.0.so.0.3000.1
0000003ba7f1e000      4K rw---  /lib64/libglib-2.0.so.0.3000.1
0000003ba7f1f000      4K rw---    [ anon ]
0000003ba8000000   1704K r-x--  /lib64/libc-2.14.90.so
0000003ba81aa000   2048K -----  /lib64/libc-2.14.90.so
0000003ba83aa000     16K r----  /lib64/libc-2.14.90.so
0000003ba83ae000      8K rw---  /lib64/libc-2.14.90.so
0000003ba83b0000     20K rw---    [ anon ]
0000003ba8400000    524K r-x--  /lib64/libm-2.14.90.so
0000003ba8483000   2044K -----  /lib64/libm-2.14.90.so
0000003ba8682000      4K r----  /lib64/libm-2.14.90.so
0000003ba8683000      4K rw---  /lib64/libm-2.14.90.so
0000003ba8800000     92K r-x--  /lib64/libpthread-2.14.90.so
0000003ba8817000   2044K -----  /lib64/libpthread-2.14.90.so
0000003ba8a16000      4K r----  /lib64/libpthread-2.14.90.so
0000003ba8a17000      4K rw---  /lib64/libpthread-2.14.90.so
0000003ba8a18000     16K rw---    [ anon ]
0000003ba9000000     28K r-x--  /lib64/librt-2.14.90.so
0000003ba9007000   2044K -----  /lib64/librt-2.14.90.so
0000003ba9206000      4K r----  /lib64/librt-2.14.90.so
0000003ba9207000      4K rw---  /lib64/librt-2.14.90.so
0000003ba9400000     84K r-x--  /lib64/libgcc_s-4.6.2-20111027.so.1
0000003ba9415000   2044K -----  /lib64/libgcc_s-4.6.2-20111027.so.1
0000003ba9614000      4K rw---  /lib64/libgcc_s-4.6.2-20111027.so.1
0000003ba9c00000     32K r-x--  /usr/lib64/libnuma.so.1
0000003ba9c08000   2048K -----  /usr/lib64/libnuma.so.1
0000003ba9e08000      4K rw---  /usr/lib64/libnuma.so.1
0000003bb6a00000     16K r-x--  /lib64/libcap-ng.so.0.0.0
0000003bb6a04000   2044K -----  /lib64/libcap-ng.so.0.0.0
0000003bb6c03000      4K r----  /lib64/libcap-ng.so.0.0.0
0000003bb6c04000      4K rw---  /lib64/libcap-ng.so.0.0.0
00007f020fc32000     24K rw---    [ anon ]
00007f020fc5e000      8K rw---    [ anon ]
00007ffff6b2b000    132K rw---    [ stack ]
00007ffff6bff000      4K r-x--    [ anon ]
ffffffffff600000      4K r-x--    [ anon ]
 total            37852K

Comment 1 Jonathan Abbey 2011-12-06 19:44:30 UTC

Saw this as well on an eight core Fedora 16 system that had been running for two weeks.  It had got up to 400 megs of RAM.

irqbalance-1.0.3-1.fc16.x86_64 is in fedora-updates-testing.repo, it should
have the fix from upstream.

Comment 2 Trevor Cordes 2011-12-10 08:19:12 UTC

"Me too".  278MB after 9 days.

Haven't tried the testing fix, but wanted to add a link to the same problem here:
http://code.google.com/p/irqbalance/issues/detail?id=22

Comment 3 Anton Arapov 2011-12-12 11:01:33 UTC

Thanks for the report, I will push an update this week.

Comment 4 Anton Arapov 2011-12-12 11:03:10 UTC

Heh, seems Neil already pushed it. And it has the patch to the reported issue.
Try the version 1.0.3-1 please.