Bug 589667

Summary: VDSM: VM stuck at "Migration Destination" status forever ( due to python deadlock )
Product: Red Hat Enterprise Linux 5 Reporter: Oded Ramraz <oramraz>
Component: pythonAssignee: Dave Malcolm <dmalcolm>
Status: CLOSED WONTFIX QA Contact: BaseOS QE - Apps <qe-baseos-apps>
Severity: high Docs Contact:
Priority: low    
Version: 5.5CC: abaron, danken, iheim, katzj, Rhev-m-bugs, smizrahi, ykaul
Target Milestone: rc   
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2013-03-21 04:37:08 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
parent process core
none
child process core none

Description Oded Ramraz 2010-05-06 16:06:11 UTC
Created attachment 412105 [details]
parent process core

Description of problem:

After running multiple migrations from host A to host B one of the guests stuck at "Migration Destination" status forever.
After running "pgrep vdsm" command on the host i noticed that there are 3 vdsm processes instead of two. 

According to danken this issue happen due to python deadlock. 


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

vdsm22-4.5-54.101870224

How reproducible:


Steps to Reproduce:
1.
2.
3.
  
Actual results:


Expected results:


Additional info:

Comment 1 Oded Ramraz 2010-05-06 16:06:55 UTC
Created attachment 412107 [details]
child process core

Comment 3 Dan Kenigsberg 2010-05-06 16:30:29 UTC
To be more exact: following the backtrace of the child process hints that it was created by subprocess module. Before the expected exec(), garbage collection fired, and vdsm's task.Task.__del__ was called. The latter blocks deep inside logging module.

It might be a bug in one of the mentioned python modules, or that logging should not be used in __del__, or something completely different.

Comment 4 Ayal Baron 2010-05-12 05:47:14 UTC
Please attach vdsm log.

Comment 5 Dave Malcolm 2010-05-24 16:34:19 UTC
Do you have a link to the .py code that's being executed?

Comment 6 Dan Kenigsberg 2010-05-25 06:43:47 UTC
http://bob.eng.lab/builds/sm64/vdsm22-4.5-54.el5rhev.x86_64.rpm should be pretty close.

Comment 8 RHEL Program Management 2011-09-23 00:34:52 UTC
This request was evaluated by Red Hat Product Management for
inclusion in the current release of Red Hat Enterprise Linux.
Because the affected component is not scheduled to be updated in the
current release, Red Hat is unfortunately unable to address this
request at this time. Red Hat invites you to ask your support
representative to propose this request, if appropriate and relevant,
in the next release of Red Hat Enterprise Linux.

Comment 10 RHEL Program Management 2013-03-21 04:37:08 UTC
Development Management has reviewed and declined this request.
You may appeal this decision by reopening this request.

Comment 11 Dan Kenigsberg 2013-03-21 07:40:59 UTC
I don't mind this bug closure, as we are using https://pypi.python.org/pypi/cpopen for quite some time in order to avoid this issue (and others).