[Pacemaker] error in logs - cannot interleave clone... and some general ideas

[Andrew Beekhof]

2010/9/15 Michał Purzyński <michal at rsbac.org>:
> hey,
>
> i'm in process of setting up a failover + load balancing cluster for
> OpenLDAP (mirror mode).
>
> i've got some strange log messages i don't really understand:
>
> Sep 15 00:33:38 udalia3 crmd: [30332]: ERROR: crmd_ha_msg_filter:
> Another DC detected: udalia4 (op=noop)
>
> this one is easy - one node lost connectivity, each of them (i have two
> node cluster) thinks he's the DC. anyway, after the connectivity is
> back, everything syncs.
>
> btw, what should i do about that? stonith?

almost certainly

> does not seem to be a good
> solution with slapd service - no shared data, each node has own
> database, they are synchronized on the ldap replication level, mirror mode.

what happens if changes are made to both?

>
> if you have some ideas/thoughts about my cluster setup, share them as
> well. maybe there are some logical mistakes i made or sth. just comment
> on it.
>
> Sep 15 00:36:53 udalia3 pengine: [30331]: ERROR:
> clone_rsc_colocation_rh: Cannot interleave clone LDAPclone and
> ClusterIPclone because they do not support the same number of resources
> per node
>
> this is the one i don't get. i don't interleave anything. at least i
> don't know i do ;)

it became the default.
set the interleave=false param for both clones and it will go away

>
> configuration of my cluster:
>
> node udalia3 \
>        attributes standby="off"
> node udalia4 \
>        attributes standby="off"
> primitive ClusterIP ocf:heartbeat:IPaddr2 \
>        params ip="172.16.6.93" cidr_netmask="24"
> clusterip_hash="sourceip-sourceport" \
>        op monitor interval="0" timeout="60s" start stop
> primitive LDAP lsb:slapd \
>        op monitor interval="0" timeout="60s" start stop
> primitive resPing ocf:pacemaker:ping \
>        params host_list="172.16.4.7" dampen="5s" multiplier="1000"
> debug="true" \
>        op start interval="0" timeout="60s" \
>        op stop interval="0" timeout="60s" \
>        op monitor interval="0" timeout="60s"
> clone ClusterIPclone ClusterIP \
>        meta globally-unique="true" clone-max="2" clone-node-max="2"
> target-role="Started"
> clone LDAPclone LDAP \
>        meta globally-unique="false" clone-max="2" clone-node-max="1"
> target-role="Started"
> clone clonePing resPing
> location ip-no-convectivity ClusterIPclone \
>        rule $id="ping-exclude-rule2" -inf: not_defined pingd or pingd
> number:lte 0
> location ldap-no-convectivity LDAPclone \
>        rule $id="ping-exclude-rule" -inf: not_defined pingd or pingd
> number:lte 0
> colocation ldap-with-ip inf: LDAPclone ClusterIPclone
> order ldap-after-ip inf: ClusterIPclone LDAPclone
> property $id="cib-bootstrap-options" \
>        dc-version="1.1.2-53b05d88305c603b83b6dbdd799f8e3bca9b7efd" \
>        cluster-infrastructure="openais" \
>        expected-quorum-votes="2" \
>        stonith-enabled="false" \
>        no-quorum-policy="ignore" \
>        last-lrm-refresh="1284502285"
> rsc_defaults $id="rsc-options" \
>        resource-stickiness="0"
>
> thx for your help.
>
> _______________________________________________
> 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://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker
>