I have a two-node cluster running NW65SP5 with shared SAN volume. The
resource being clustered is NDPS manager and broker. When a node
abends, the Master IP Address Resolver fails over to the other node, but
the cluster server resource goes comatose and must be taken offline
then migrated.

Here's my cluster load script:
nss /poolactivate=NDPSC1P1
mount NDPSC1VOL1 VOLID=254
search add NDPSC1VOL1:system
CLUSTER CVSBIND ADD NWNDPSVS1 xx.xx.xx.xx
NUDP ADD NWNDPSVS1 xx.xx.xx.xx
add secondary ipaddress xx.xx.xx.xx
NSS /Transaction=ndpsc1vol1
ap2webdn
ap2webup
load broker .ndpsbroker1.ndps.campus.org /allowdup /ipaddress=10.127.16.165
load ndpsm .iprint1.ndps.campus.org /dbvolume=nocheck /ipaddress=xx.xx.xx.xx
search add ndpsc1vol1:system
search add ndpsc1vol1:system\tivoli\tsm\client\ba
tsmu

---------------------------------------------------

Thoughts?