[Pacemaker] Is there any way to reduce the time for migration of the resource from one node to another node in a cluster on failover.

rakesh k rakirocker4236 at gmail.com
Tue Mar 22 07:41:51 EDT 2011


Hi All

I am providing you the configuration I used for testing the resource
migration.

   - Node-1 resource failed .
   - Message sent to node-2
   - the log message i found in ha-debug file (pengine: [15991]: notice:
   common_apply_stickiness: Tomcat1 can fail 999999 more times on mysql3 before
   being forced off)
   - The process is getting started on same node where it fails after time
   out.
   - stopping virtual IP
   - send note to second node-2
   - Starting VIP
   - second node starts the process.

The total duration it is taking is about *one and half minute *is there any
way to reduce the time for this sceanrio.

Plese find the configuration i used

node $id="6317f856-e57b-4a03-acf1-ca81af4f19ce" cisco-demomsf
node $id="87b8b88e-3ded-4e34-8708-46f7afe62935" mysql3
primitive Tomcat1 ocf:heartbeat:tomcat \
        params tomcat_name="tomcat" statusurl="
http://localhost:8080/dbtest/testtomcat.html" java_home="/"
catalina_home="/home/msf/runtime/tomcat/apache-tomcat-6.0.18" client="curl"
testregex="*</html>" \
        op start interval="0" timeout="60s" \
        op monitor interval="50s" timeout="50s" \
        op stop interval="0" \
        meta target-role="Started"
primitive Tomcat1VIP ocf:heartbeat:IPaddr3 \
        params ip="<ip address>" eth_num="eth0:2"
vip_cleanup_file="/var/run/bigha.pid" \
        op start interval="0" timeout="120s" \
        op monitor interval="30s" \
        meta target-role="Started"
colocation Tomcat1-with-ip inf: Tomcat1VIP Tomcat1
order Tomcat1-after-ip inf: Tomcat1VIP Tomcat1
property $id="cib-bootstrap-options" \
        dc-version="1.0.9-89bd754939df5150de7cd76835f98fe90851b677" \
        cluster-infrastructure="Heartbeat" \
        stonith-enabled="false" \
        no-quorum-policy="ignore" \
        last-lrm-refresh="1300787402"
rsc_defaults $id="rsc-options" \
        resource-stickiness="500"
Regards
Rakesh
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.clusterlabs.org/pipermail/pacemaker/attachments/20110322/06f750ae/attachment-0002.html>


More information about the Pacemaker mailing list