[Pacemaker] [ClusterLabs] Unexpected behaviour of PEngine Recheck Timer while in maintenance mode

Andrew Beekhof andrew at beekhof.net
Mon May 18 02:34:17 UTC 2015


> On 29 Apr 2015, at 5:40 am, Rolf Weber <Rolf.Weber at asamnet.de> wrote:
> 
> Hi!
> 
> On 07:27 Mon 27 Apr     , Andrew Beekhof wrote:
>> What exactly were you doing at this point?
> 
> resizing a filesystem.
> fs was unexported and unmounted.
> as I understand maintenence mode this should have worked (and there was no
> problem until the recheck timer triggered).
> 
>> I ask because:
>> 
>> Apr 19 23:16:51 astorage1 crmd: [4150]: ERROR: lrm_get_rsc(666): failed to send a getrsc message to lrmd via ch_cmd channel.
>> Apr 19 23:16:51 astorage1 crmd: [4150]: ERROR: lrm_get_rsc(666): failed to send a getrsc message to lrmd via ch_cmd channel.
>> Apr 19 23:16:51 astorage1 crmd: [4150]: ERROR: lrm_add_rsc(870): failed to send a addrsc message to lrmd via ch_cmd channel.
>> Apr 19 23:16:51 astorage1 crmd: [4150]: ERROR: lrm_get_rsc(666): failed to send a getrsc message to lrmd via ch_cmd channel.
>> 
>> and 
>> 
>> Apr 19 23:16:52 astorage2 crmd: [4018]: CRIT: lrm_connection_destroy: LRM Connection failed
>> Apr 19 23:16:52 astorage2 crmd: [4018]: info: lrm_connection_destroy: LRM Connection disconnected
>> 
>> suggest that the lrmd processes on both machines crashed or failed.
> 
> I know. I didn't find any core files or other indications that it indeed did
> crash.

if you were resizing the filesystem that pacemaker was running on, then i’m not surprised (either that the processes aborted or that there was no core file).

> errors appeared after the failed-message for op 43 (which is ipmi fencing).
> 
>> I would definitely recommend an upgrade from 1.1.7
> 
> will be done, now that jessie is stable. don't quite know when yet.
> 
> cheers,
> rolf weber
> 
> 
> _______________________________________________
> Users mailing list: Users at clusterlabs.org
> 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





More information about the Pacemaker mailing list