[Pacemaker] Monitoring on master node not running after standby is connected

Andrew Beekhof andrew at beekhof.net
Mon Sep 23 05:53:14 UTC 2013


On 20/09/2013, at 5:45 AM, Juraj Fabo <juraj.fabo at gmail.com> wrote:

> Juraj Fabo <juraj.fabo at ...> writes:
> 
>> 
>> Dear all
>> 
>> Attached is my 2-nodes, master slave cluster configuration with master-slave
>> postgresql resource and some IP resources.
>> I've modified pgsql resource agent to log its "main" entry with the
>> parameter to see what operation is called.
>> My problem is that while the single node is running, the monitor operation
>> on pgsql resource is running.
>> Even after promoting to master.
>> 
>> However, when second node is connected, monitor operation is no longer
>> executed on master node, only on hot-standby.
>> I've seen very similar bug report
>> http://bugs.clusterlabs.org/show_bug.cgi?id=5072 which was quite some time
>> ago fixed.
>> I know that the monitor operation on MS resource must be configured twice
>> with different intervals to work properly 
>> according to
>> 
> 
> 
> Hi
> 
> in verbose logs I found messages that the pgsql resource agent on running
> master is notified with pre/post stop operation when second node connects
> and the monitoring operation on master is cancelled.
> 
> in my log "ifds1" is master and "ifds2" slave (or slave candidate)
> 
> but why is the pgsql RA on master monitoring cancelled?

Really hard to say based only on this.
Can you create a crm_report archive that starts before ifds2 comes online?

> 
> 
> Juraj
> 
> Sep 19 21:25:41 [17141] ifds1       crmd:    debug: run_graph:  Transition 7
> (Complete=0, Pending=2, Fired=4, Skipped=0, Incomplete=10,
> Source=/var/lib/pacemaker/pengine/pe-input-135.bz2): In-progr
> ess
> Sep 19 21:25:41 [17141] ifds1       crmd:   notice: te_rsc_command:    
> Initiating action 64: notify IFDS-pgsql_pre_notify_stop_0 on ifds1 (local)
> Sep 19 21:25:41 [17138] ifds1       lrmd:    debug: process_lrmd_message:  
>    Processed lrmd_rsc_info operation from
> a2d8bec8-8e67-4b76-b576-ad9915deae20: rc=0, reply=0, notify=0, exit=4202544
> Sep 19 21:25:41 [17141] ifds1       crmd:    debug:
> stop_recurring_action_by_rsc:       Cancelling op 73 for IFDS-pgsql
> (IFDS-pgsql:73)
> Sep 19 21:25:41 [17141] ifds1       crmd:    debug: cancel_op:  Cancelling
> op 73 for IFDS-pgsql (IFDS-pgsql:73)
> Sep 19 21:25:41 [17138] ifds1       lrmd:     info: cancel_recurring_action:
>   Cancelling operation IFDS-pgsql_monitor_7000
> Sep 19 21:25:41 [17138] ifds1       lrmd:    debug: log_finished:      
> finished - rsc:IFDS-pgsql action:monitor call_id:73  exit-code:8
> exec-time:0ms queue-time:0ms
> Sep 19 21:25:41 [17138] ifds1       lrmd:    debug: process_lrmd_message:  
>    Processed lrmd_rsc_cancel operation from
> a2d8bec8-8e67-4b76-b576-ad9915deae20: rc=0, reply=1, notify=0, exit=4202544
> Sep 19 21:25:41 [17141] ifds1       crmd:    debug: cancel_op:  Op 73 for
> IFDS-pgsql (IFDS-pgsql:73): cancelled
> ..
> Sep 19 21:25:41 [17141] ifds1       crmd:   notice: te_rsc_command:    
> Initiating action 64: notify IFDS-pgsql_pre_notify_stop_0 on ifds1 (local)
> 
> 
> 
> 
> 
> 
> _______________________________________________
> 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/20130923/a44e12b1/attachment-0009.sig>


More information about the Pacemaker mailing list