>>> Ken Gaillot <[email protected]> schrieb am 09.11.2016 um 16:43 in 
>>> Nachricht
<[email protected]>:
> On 11/09/2016 02:33 AM, Ulrich Windl wrote:
>>>>> Ryan Anstey <[email protected]> schrieb am 08.11.2016 um 19:54 in 
> Nachricht
>>> Log when running cleaning up the resource on the OLD node:
>>>
>>> Nov 08 09:21:18 h3 crmd[11394]: warning: No match for shutdown action on
>>> 167838209
>> 
>> This indicates a node communication problem!
> 
> Not necessarily; it's fairly routine. In 1.1.15, we changed the wording
> to "No reason to expect node to be down", and in 1.1.16, we only log it
> if there's a problem.

I only know the message when a node was shut down by means separate from 
pacemaker, or when the communication was disturbed.

> 
> It indicates that Pacemaker checked whether a node was expected to be
> down. It does that routinely for certain normal events to react
> appropriately, so most of the time it's unimportant. It's only a problem
> if the node went down, so since 1.1.16, it's only logged in that situation.

In every case I've seen, pacemaker scheduled a STONITH for that node next.
(That made me think pacemeaker thinks the node is down and now wants to make 
sure it actually is)

> 
> _______________________________________________
> Users mailing list: [email protected] 
> http://clusterlabs.org/mailman/listinfo/users 
> 
> Project Home: http://www.clusterlabs.org 
> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf 
> Bugs: http://bugs.clusterlabs.org 





_______________________________________________
Users mailing list: [email protected]
http://clusterlabs.org/mailman/listinfo/users

Project Home: http://www.clusterlabs.org
Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
Bugs: http://bugs.clusterlabs.org

Reply via email to