Bug 1044415

Summary: Libvirtd CPU usage goes to 100%
Product: Red Hat Enterprise Linux 6 Reporter: pagupta
Component: kernelAssignee: Dave Anderson <anderson>
kernel sub component: Virtualization QA Contact: Virtualization Bugs <virt-bugs>
Status: CLOSED WONTFIX Docs Contact:
Severity: urgent    
Priority: urgent CC: dyuan, honzhang, jdenemar, jraju, mprivozn, mzhan, pagupta, rbalakri, riehecky, rpai, sankarshan, yanwang, zhwang
Version: 6.4   
Target Milestone: rc   
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2015-10-14 16:03:03 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:    
Bug Blocks: 1075802, 1159933, 1172231, 1270638    
Attachments:
Description Flags
gdb core of libvirt none

Description pagupta 2013-12-18 09:36:25 UTC 
Description of problem:

Libvirtd CPU usage goes to 100% for more than 15 minutes.

Customer tried to get core dump but could not attach gdb to libvirtd as it was not responding during the time of issue.


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

libvirt-0.10.2-18.el6_4.14.x86_64


How reproducible:

Steps to Reproduce:
1.
2.
3.

Actual results: libvirtd CPU usage goes to 100% for more than 15 minutes.


Expected results:


Additional info:
Comment 9 Jaison Raju 2014-02-21 12:18:07 UTC 
Created attachment 865954 [details]
gdb core of libvirt

libvirt-0.10.2-29.el6.1.x86_64
Comment 10 Michal Privoznik 2014-02-28 16:13:07 UTC 
(In reply to Jaison Raju from comment #9)
> Created attachment 865954 [details]
> gdb core of libvirt
> 
> libvirt-0.10.2-29.el6.1.x86_64

I don't think I can read much from the core file. I mean, a core file is created in a single shot hence it captures only one state that an application is currently in. And this can be completely unrelated state. In order to identify bottleneck I need to get a consecutive states = profile info. The best would be to gather profile data via oprofile which offers insight into whole system, not just a single application. Jaison, can you try to get the profile data?