[Pacemaker] Please Help - frequent cleanup is required for the resources on failover condition

Abhin.G.S - DEUCN deucn at inMail.sk
Tue Aug 18 11:40:30 UTC 2009


Hello Dejan,

Thank you for the walk through.

Special thanks to lmb, andrew.

Have fun,

wish you a wonderful time.

regards,

DEUCN Advanced Computing Lab


---- Original message ----
From: Dejan Muhamedagic <dejanmm at fastmail.fm>
To: pacemaker at oss.clusterlabs.org
Date: 8/18/2009 1:43:00 PM
Subject: Re: [Pacemaker] Please Help - frequent cleanup is required for the resources on failover condition 

Hi,  On Tue, Aug 18, 2009 at 07:08:16AM +0530, Abhin GS wrote: > Good Morning Dejan, >  > Thank you for your input, will increase the time out to 60 sec. I guess > we have switched the other machine off (pull the plug we did), the > fencing was strapped to "ilore", i guess since there was no supply to > node2, HP ilo2 did not function, so node 1 could not reach it.  >  > Pull the plug with ilo as the fencing method was a dump thing to do i > guess.  >  > So in a scenario where the node1 cannot reach the other fencing device > will break the cluster?     With ilo devices yes.  > extract of a mail form the group. >  > # On Mon, Aug 17, 2009 at 04:18:34PM -0600, hj lee wrote: > #> Hi, > #>  > #> I defined stonith:ssh, they are running in both machines as a clone. > #How is > # > #How do you expect ssh to work if you pull the power plug from the > #target node? > # > #> the stonith related to promoting standby? When the Pacemaker detects > #master > #> node was gone in cluster, then why doesn't Pacemaker promote the > #standby? > # > #In case of such an event, when a node disappears without saying > #goodbye, the other node has to make sure that the node which left > #is down and not only unreachable, that's where fencing comes in. > # > #Thanks, > # > #Dejan >  > I guess the answer for my question is "yes". Correct me if i am wrong. >  > If this is the case, for some reason, if i had to replace the node 2 > when node1 is dc, could break the cluster right, coz the fencing device > will be unreachable.  In that case you can put the node in standby before removing it from the cluster. Or just shut it down. That's an entirely different case, because the node will leave the cluster in an orderly manner.  Thanks,  Dejan  > Please do let me know if there were any another anomalies. >  >  > Thank You, >  > Warm Regards >  > Abhin >  > On Mon, 2009-08-17 at 13:35 +0200, Dejan Muhamedagic wrote: > > Dejan >  > ---------- > Ukazte svoje fotky na www.zonerama.sk  _______________________________________________ Pacemaker mailing list Pacemaker at oss.clusterlabs.org http://oss.clusterlabs.org/mailman/listinfo/pacemaker

------------------------------------
Abhin.G.S
=========
+91-9895-525880  |  +91-471-2437189
D E U C N ®  | http://www.deucn.com
------------------------------------

----------

Ukazte svoje fotky na www.zonerama.sk
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20090818/a9939609/attachment-0002.htm>


More information about the Pacemaker mailing list