Hi All!

Ok dangit, I was up for 40 hours when I wrote my last post. Please
delete that one (Subject: NW65SP3 NWCS 1.7.2 One way only failover only)
and I'll fix the errors in this one.


The issue:
I have a 2 node cluster that the volume resource only failsover from 2
to 1, never from 1 to 2. Identical servers (IBM 345), QL2300.ham 6.80.03.
On a cold reboot of the server1 (the one that it can't failover to) it
mounts fine if not already mounted on server2. It also mounts correctly
when manualy migrated to either server. The end result is the resource
goes comotose and can be manually offined and onlined successfully.


clstrlib /hmo=off [set in ldncs.ncf on both]


Scripts:
Load:

set allow ip address duplicates=on
nss /poolactivate=SHARK
mount DATA VOLID=254
CLUSTER CVSBIND ADD APPSVS 172.17.10.96
NUDP ADD APPSVS 172.17.10.96
add secondary ipaddress 172.17.10.96
load broker .APPS-cl-broker.DEV /allowdup /ipaddress=172.17.10.96
CIFS ADD .CN=APPSVS.O=DEV.T=DEV-TREE.
load ndpsm .ndps-manager.DEV /dbvolume=nocheck /ipaddress=172.17.10.96
set allow ip address duplicates=off

Unload:

unload ndpsm
unload broker
Delay 15
del secondary ipaddress 172.17.10.96
NUDP DEL APPSVS 172.17.10.96
CLUSTER CVSBIND DEL APPSVS 172.17.10.96
CIFS DEL .CN=APPSVS.O=DEV.T=DEV-TREE.
nss /pooldeactivate=SHARK /overridetype=question


Thanks for any thoughts!

Joe