[Pacemaker] Configuration question

[u.schmeling at online.de]

An HTML attachment was scrubbed...
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20110428/23ad4da6/attachment-0003.html>
-------------- next part --------------

Hi List, http://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker 

my configuration consists of a 2 node cluster running a haproxy/apache environment. The active apache/haproxy config is bound to a pair of virtual addresses haproxyIP/apacheIP. There is a resource "webservice" which is monitoring the active apache/haproxy configuration. On the passive node I would like to have a webserviceStandby running, which is monitoring the passive (running) haproxy/apache configuration. In case of failure, the standby node should be set to standby to prevent a migration of the active node to a non working passive node. This is my configuration:

node mydpd-lbv01.dpd.de \
 attributes standby="off"
node mydpd-lbv02.dpd.de \
 attributes standby="off"
primitive apacheIP ocf:heartbeat:IPaddr2 \
 params ip="10.6.151.190" \
 op monitor interval="10s" \
 meta is-managed="true" target-role="Started"
primitive haproxyIP ocf:heartbeat:IPaddr2 \
 params ip="10.6.151.191" \
 op monitor interval="10s" \
 meta target-role="Started"
primitive notifier ocf:heartbeat:MailTo \
 params email="uwe.schmeling at ubl-is.de"
primitive pingd ocf:pacemaker:ping \
 params host_list="10.6.151.11" multiplier="100" \
 op monitor interval="15s" timeout="5s"
primitive webservice ocf:myagents:webservices \
 params role="active" \
 op monitor on-fail="standby" interval="15s" \
 meta failure-timeout="20s" migration-threshold="2" is-managed="true" target-role="Started"
primitive webserviceStandby ocf:myagents:webservices \
 op monitor interval="15s" timeout="10s" on-fail="standby" \
 params role="standby" \
 meta target-role="Started" is-managed="true" failure-timeout="20s"
group webservice-ips haproxyIP apacheIP notifier \
 meta target-role="Started"
colocation all-webresources-on-one-node inf: haproxyIP apacheIP
colocation standbyrestriction -inf: webservice-ips webserviceStandby
colocation webservice-and-ip inf: haproxyIP apacheIP webservice
order webservice-before-standby inf: webservice webserviceStandby
order webservice-ips-before-webservice inf: webservice-ips webservice
property $id="cib-bootstrap-options" \
 dc-version="1.1.2-f059ec7ced7a86f18e5490b67ebf4a0b963bccfe" \
 cluster-infrastructure="openais" \
 expected-quorum-votes="2" \
 stonith-enabled="false" \
 no-quorum-policy="ignore" \
 last-lrm-refresh="1303993830" \
 cluster-delay="30"
rsc_defaults $id="rsc-options" \
 resource-stickiness="100"

If I set one node to standby, the cluster decides to run the webserviceStandby on the remaining node and stops all other resources. This is exactly the opposite of what I want to see. How would I modify the configuration, to give the webservice & co priority over webserviceStandby. Any help would be appreciated....

regards Uwe