[Pacemaker] pacemaker with pgpool2

[Andrew Beekhof]

On Sun, May 30, 2010 at 4:37 PM, Démi Zsolt <zsdemi at gmail.com> wrote:
> Hello,
>
> I have a problem to control pgpool2 from pacemaker crm shell. The pgpool2
> init script is working correctly from command line, but the lsb resource
> not.
>
> Here is my config :
> node $id="144a904b-6a02-477c-ae2e-22d4c6a6f085" server-lb1 \
>    attributes standby="off"
> node $id="68f7b8d4-5921-4f5b-b77b-21a3b18909b3" server-lb2 \
>    attributes standby="off"
> primitive kulso_virtual ocf:heartbeat:IPaddr2 \
>    params ip="192.168.2.13" nic="eth0" cidr_netmask="24" \
>    op monitor interval="10s" timeout="20s" depth="0" \
>    meta target-role="Started"
> primitive pgpool_srv lsb:pgpool2 \
>    op monitor interval="10s" timeout="20s" \
>    meta target-role="Stopped"
> property $id="cib-bootstrap-options" \
>    dc-version="1.0.8-042548a451fce8400660f6031f4da6f0223dd5dd" \
>    cluster-infrastructure="Heartbeat" \
>    is-managed-default="true" \
>    stonith-enabled="false" \
>    last-lrm-refresh="1275228322"
>
> Here is the error messages what I see in the crm shell :
> Stack: Heartbeat
> Current DC: server-lb2 (68f7b8d4-5921-4f5b-b77b-21a3b18909b3) - partition
> with quorum
> Version: 1.0.8-042548a451fce8400660f6031f4da6f0223dd5dd
> 2 Nodes configured, unknown expected votes
> 2 Resources configured.
> ============
>
> Online: [ server-lb1 server-lb2 ]
>
> kulso_virtual   (ocf::heartbeat:IPaddr2): [7CStarted server-lb1
> pgpool_srv [6C(lsb:pgpool) Started [  server-lb2 [6Cserver-lb1 ]
>
> Failed actions:
>    pgpool_srv_monitor_0 (node=server-lb2, call=8, rc=1, status=complete):
> unknown error
>    pgpool_srv_stop_0 (node=server-lb2, call=9, rc=1, status=complete):
> unknown error
>    pgpool_srv_monitor_0 (node=server-lb1, call=6, rc=1, status=complete):
> unknown error
>    pgpool_srv_stop_0 (node=server-lb1, call=7, rc=1, status=complete):
> unknown error
>
>
> The pgpool not running on any node. How CRM monitor the resource?

When Pacemaker starts, before it does _anything_, it asks the new node
if it already has anything running (with a non-recurring monitor op).

> Why crm
> think's it's running?

The monitor op failed, so the only safe thing to do is assume it is running.
Then your stop action failed (which it must not if the resource is
truly stopped).
And since fencing is disabled, it couldn't make any progress.

So:
1) Use ocf-tester to fix your agent because its not currently ocf compliant
2) Enable and configure stonith

> Unfortunately I can't see anything related in the
> logs. Can I increase the pacemaker logging level?
> Thank you for any help in advance!
>
> Best Regards
>    Zsolt
>
> _______________________________________________
> 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
>