Assuming that your cluster services are stopped and the resource groups are running as UNMANEGED.
- Start cluster services using setting "Manage Resource Groups" to "Manually" on the node were you stopped the services and let the resource groups as unmaneged
- Create dummy monitoring script
cat > /tmp/dummy.sh
#!/bin/sh
exit 0
chmod +x /tmp/dummy.sh
- List cluster's applicatios
cllsserv
- Add the dummy monitoring script as the monitoring script for the application (testapp_rg1 on this example)
- Note 1: You have to do this for all the applications managed by your clsuter.
- Note 2: You must use a single "name" for each application monitoring script.
/usr/es/sbin/cluster/utilities/claddappmon MONITOR_TYPE=user name='dummymon' RESOURCE_TO_MONITOR='testapp_rg1' INVOCATION='both' MONITOR_METHOD='/tmp/dummy.sh' STABILIZATION_INTERVAL='5' RESTART_COUNT='5' FAILURE_ACTION='notify'
- Command to generate all necessary claddappmon commands. Run the commands manually after you run below command.
for RESOURCE in $(cllsserv | awk '{print $1}') ; do echo "/usr/es/sbin/cluster/utilities/claddappmon MONITOR_TYPE=user name='$RESOURCE.dummymon' RESOURCE_TO_MONITOR='$RESOURCE' INVOCATION='both' MONITOR_METHOD='/tmp/dummy.sh' STABILIZATION_INTERVAL='5' RESTART_COUNT='5' FAILURE_ACTION='notify'" ; done
- Check if the monitoring script has been added
/usr/es/sbin/cluster/utilities/cllsappmon
- Synchronize the cluster
/usr/es/sbin/cluster/utilities/cldare -rt -V 'normal' -C'interactive'
- Check return code for the monitoring script, it has to be 0 or HACMP will rerun the application start scripts
/usr/es/sbin/cluster/utilities/cl_app_startup_monitor -s testapp_rg1
echo $?
- Bring Resource Online via CSPOC
- Check cluster status
- Remove the monitoring script from cluster configuration and synchronize it
/usr/es/sbin/cluster/utilities/clrmappmon dummymon
/usr/es/sbin/cluster/utilities/cllsappmon
/usr/es/sbin/cluster/utilities/cldare -rt -V 'normal' -C'interactive'