Bug 608500

Summary:	fence_ilo intermittently leaves server offline.
Product:	[Retired] Red Hat Cluster Suite	Reporter:	Jason Nelson <jason.nelson>
Component:	fence	Assignee:	Marek Grac <mgrac>
Status:	CLOSED DUPLICATE	QA Contact:	Cluster QE <mspqa-list>
Severity:	medium	Docs Contact:
Priority:	low
Version:	4	CC:	cluster-maint, edamato
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:	2010-07-14 12:54:33 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:

Description Jason Nelson 2010-06-27 21:18:07 UTC

Description of problem:

When fence_ilo fires off, the server will power off, but will remain in a power off state

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

This has been seen as recently as the fenced deployed with 5.4


How reproducible:

We're seeing this on DL385, and DL385G2.  Have not yet tested on the DL580s we have.


Steps to Reproduce:
1.  Setup working cluster
2.  Kernel panic active node
3.  Fence will power off panic'd node, but will not power back on.
  
Actual results:

System is left with power off.  Can be verified via telnet/ssh into iLo, and viewing power status.

Expected results:

Server should power back on after power is turned off.

Additional info:

We've been resolving this on our end by editing the /sbin/fence_ilo script to do:

poweroff
poweron
poweron

Running poweron twice has fixed it, I have not tested by putting a delay in between off and on.

BTW, nice meeting the team at Summit.  Like I said, we'll try to keep you in the loop on the bugs we see that we workaround in RHCS.

-Jason Nelson
Lead Linux Engineer
Rackspace Hosting

Comment 1 Marek Grac 2010-06-28 12:44:53 UTC

Can you take a look at bug #545682 ? I believe that it is same problem

Comment 2 Jason Nelson 2010-06-28 20:03:00 UTC

Ah, does look like a duplicate here.  Is this going to be back ported into earlier versions of RHEL 4/5?

-Jason Nelson
Lead Linux Engineer
Rackspace Hosting

Comment 3 Marek Grac 2010-07-14 12:54:33 UTC

@Jason:

Do you mean?
* 4.9 update (yes)
* z-stream (yes, already in 4.8.z)
* in RHEL5/6 (yes, not cloned from same bugs as timing options were added for all fence agents)

*** This bug has been marked as a duplicate of bug 545682 ***