[Pacemaker] problem with resource group

[Andrew Beekhof]

On Sat, Jun 9, 2012 at 5:49 PM, Luca Meron <ckslxpcjus at hotmail.it> wrote:
>
>
> Hi.
> I've created a 2 node active/passive cluster. The HA manages 17 resources, among IP and other services.
> But I've a problem when resource placement: after I add one of the latest resource, it is started on node2 instead of node1 with all the others!

Did you put it in the group or just add it to the configuration?
Did you add any colocation constraints for them?

>
> I couldn't really find a reason for this into syslog. I was adding a couple of resources at a time to the cluster, and when I added Cluster-winbind it was started on host2! Why is this happening?
> Also, I tried stopping corosync on host2 and all resources were started correctly on host1, but when I retarted it on host2 while primitives were started on host1 it attempted to start them on host2 also!
> I'm using ubuntu 10.04 fully updated, corosync from ubuntu-ha-maintainers ppa, corosync 1.4.2-1ubuntu0~ppa1, pacemaker 1.1.6-2ubuntu0~ppa2.
> Thanks
>
> Some parts of the config:
> primitive Cluster-FS-DRBD ocf:linbit:drbd \     params drbd_resource="r0" \     meta target-role="Started"primitive Cluster-FS-Mount ocf:heartbeat:Filesystem \         params device="/dev/drbd/by-res/r0" directory="/srv" fstype="ext3" \    meta target-role="Started"primitive Cluster-IP ocf:heartbeat:IPaddr \   params ip="192.168.1.1" cidr_netmask="24" nic="eth0" \  op monitor interval="90s" \     meta target-role="Started"[...]group rs Cluster-FS-Mount Cluster-Zimbra-FS Cluster-IP Cluster-IP2 Cluster-IP3 Cluster-nfs Cluster-smbd Cluster-nmbd Cluster-apache2 Cluster-vsftpd Cluster-winbind cron_backup cron_set1 cron_set2ms MS-r0 Cluster-FS-DRBD \    meta master-max="1" notify="true"colocation Mount-on-r0 inf: rs MS-r0:Masterorder group-after-mount inf: MS-r0:promote rs:startproperty $id="cib-bootstrap-options" \   dc-version="1.1.6-9971ebba4494012a93c03b40a2c58ec0eb60f50c" \   cluster-infrastructure="openais" \      stonith-enabled="false" \       expected-quorum-votes="2" \     last-lrm-refresh="1336942125" \         no-quorum-policy="ignore"
> _______________________________________________
> 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