[Pacemaker] do_state_transition: Starting PEngine Recheck Timer

[Andrew Beekhof]

On 10/10/2013, at 4:12 PM, Xiaomin Zhang <zhangxiaomin at gmail.com> wrote:

> Hi, Gurus:
> Do you ever see the message:
> do_state_transition: Starting PEngine Recheck Timer
> occurring when doing fail over?
> I found that while PEngine is starting the rechecker timer, the cluster seems split brain, that all nodes are not seeing each other.

The two are entirely unrelated.
The purpose of the recheck timer is to make sure any time-based rules/options take effect.
When it expires, the cib is queried and the policy engine checks if everything is still in the right spot.

There is no way it can impact the ability of the nodes to communicate.

> Although the living nodes are operating properly. 

I very much doubt that if they can't see each other.

> And this situation usually keeps for about 15 minutes to recover.
> Is this an expected behavior?
> Thanks.
> I  
> _______________________________________________
> Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
> 
> Project Home: http://www.clusterlabs.org
> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
> Bugs: http://bugs.clusterlabs.org

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 841 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20131016/75f29739/attachment-0004.sig>