Hello all,

I have a two node NetWare 6.5.6 cluster connected to a three node iSCSI SAN cluster. What I am seeing is that if the SAN module that is the active virtual IP holder gets rebooted or otherwise goes offline I get a castout abend on the NetWare cluster. It only takes the remaining SAN modules roughly 15-30 seconds to reconfigure with a new VIP holder, but that seems to be enough time to cause a problem with NetWare.

Is there some setting I can change to modify this behavior? I've tried increasing the cluster protocol fields: tolerance, heartbeat, etc. but that just seemed increase the time before the cluster failed over. Is there some way to increase the SBD timers?