We have a 2 node NW 65 SP2 cluster using SAN with 5 volumes. Sometimes
the second server, which is assigned 2 volumes, stops communicating,
receives a poison pill and the resources (two volumes), instead of being
failed over to the other server are comatose.
The log says it were quarantined in order to prevent abending the other
server.. After that, I put the resources off-line and then on-line again
and they are OK.

We are investigating the abends cause, but, in the meanwhile, is there a
way to change that behaviour (not to get the resources quarantined after
an abend) ??