We have a 3-nodes clustering environment running on Novell 6.0 connecting to HP MSA1000.

Today all the 3 nodes abend with the following error message.

CLUSTER: Node castout
fatal SAN device alert

I tried to restart the nodes and apparently the servers can be up and connected to the SAN for about 25-30 mins before they abend.

Any one know what could be the cause of the problem. Please help!

Poh Gek