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'

Topics