My AIX group has this issue...seems that something isn't configured right..

the storapid and storwatchd (EMC solutions enabler processes) processes which will put a lock on hdisks that are part of the powerpath configuration.  Since the affected powerpath devices are configured to volume groups needed by the cluster resource group, the cluster service will start, but it will not come online for the primary cluster node.   The processes are manually aborted, which will free up the affected disks, make them available for their controlling volume group to come online, and allow the cluster service to join the node into the cluster

Scott'