Bug 697656

Summary:	Trying to stop cman before rgmanager causes rgmanager to fail to stop properly.
Product:	Red Hat Enterprise Linux 6	Reporter:	Madison Kelly <mkelly>
Component:	rgmanager	Assignee:	Ryan McCabe <rmccabe>
Status:	CLOSED DUPLICATE	QA Contact:	Cluster QE <mspqa-list>
Severity:	medium	Docs Contact:
Priority:	medium
Version:	6.3	CC:	cluster-maint, edamato, mjuricek
Target Milestone:	rc
Target Release:	6.4
Hardware:	x86_64
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2012-10-09 16:27:20 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 Madison Kelly 2011-04-18 20:52:28 UTC

Description of problem:

If you disabled all services using 'clusvcadm', neglect to stop rgmanager and then try to stop cman, cman will fail to stop as expected. However, if you then try to stop rgmanager, it will fail to remove it's PID file. Starting and stopping rgmanager solves the problem and cman can then stop properly. 

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

rgmanager-2.0.52

How reproducible:

100%

Steps to Reproduce:
1. Stop all rgmanager services using clusvcadm
2. stop cman, it will fail.
3. stop rgmanager. It will *appear* to stop, but if you then check it's status, it will report that the PID exists still.
4. stop cman again, it will still fail.
5. start and stop rgmanager, and this time it will stop cleanly.
6. cman will now stop properly.
  
Actual results:

rgmanager leaves it's PID file behind, blocking cman from stopping.

Expected results:

cman should stop rgmanager *or* stopping rgmanager's init script should properly clear it's PID file in a scenario like this.

Additional info:

Comment 3 RHEL Program Management 2012-07-10 08:26:49 UTC

This request was not resolved in time for the current release.
Red Hat invites you to ask your support representative to
propose this request, if still desired, for consideration in
the next release of Red Hat Enterprise Linux.

Comment 4 RHEL Program Management 2012-07-11 01:49:13 UTC

This request was erroneously removed from consideration in Red Hat Enterprise Linux 6.4, which is currently under development.  This request will be evaluated for inclusion in Red Hat Enterprise Linux 6.4.

Comment 7 Ryan McCabe 2012-10-09 16:18:41 UTC

I can't reproduce this on RHEL6. Should this have been filed against RHEL5? The package version you noted is a RHEL5 rgmanager package. It sounds a lot like Bug 769730 which was fixed for RHEL 5.9.

When I try to reproduce this on RHEL 6.3, I see:

# service cman status
cluster is running.
# service rgmanager status
rgmanager (pid  11786) is running...
# service cman stop
Stopping cluster: 
   Leaving fence domain... found dlm lockspace /sys/kernel/dlm/rgmanager
fence_tool: cannot leave due to active systems
                                                           [FAILED]
# service cman status
cluster is running.
# service rgmanager stop
Stopping Cluster Service Manager:                          [  OK  ]
# service rgmanager status
rgmanager is stopped
# service cman stop
Stopping cluster: 
   Leaving fence domain...                                 [  OK  ]
   Stopping gfs_controld...                                [  OK  ]
   Stopping dlm_controld...                                [  OK  ]
   Stopping fenced...                                      [  OK  ]
   Stopping cman...                                        [  OK  ]
   Waiting for corosync to shutdown:                       [  OK  ]
   Unloading kernel modules...                             [  OK  ]
   Unmounting configfs...                                  [  OK  ]
# service cman status
corosync is stopped

Comment 8 Madison Kelly 2012-10-09 16:21:39 UTC

Hrm, given that it's such an old ticket and that the rgmanager version is 2, yes, probably rhel 5. I've not tested / run into this issue in recent memory, so as far as I am concerned, it's closed. You might want to give it a test against rhel5 once, just to be sure. I'll leave this to your discretion.

Cheers

Comment 9 Ryan McCabe 2012-10-09 16:27:20 UTC

Ok, I'm going to close this as a duplicate of the RHEL5 bug. That one was tested and verified by QE within the past month or so.

*** This bug has been marked as a duplicate of bug 769730 ***