We are running a 5 node cluster (2 nodes NW6.5sp7, 3 nodes OES2sp2) - moving away from NetWare. Twice with the Master_IP on an OES2sp2 server I had the server fail and the Master_IP go into a comatose state on the failed OES2sp2 server and also when it rolled to another OES2sp2 node it also went into a comatose state because it thought is still running on the failed server. The only way I found to recover was to shut everything down and reboot all the cluster nodes at the same time.

Is there an additional setting I should add to the OES2sp2 server configuration so the server does a hard ABEND like on the NetWare servers? I notice that an OES2 server will stop working but not really ABEND or take the poison pill and leave the cluster. I think what is happening is the OES2sp2 server fails epoch and just goes into a suspend mode not forcing the resources on that node to roll to other nodes.

Thanks for any advice.