Whilst I've figured out how to get this to work (migrating from NW 6.5.8 NCS to a rolling cluster OES2 SP2 Linux NCS), I have a few questions:

1) When I "configure" the agents on the OES2 Linux node, I figured out that I do NOT want to configure them for clustering (apparently that assumes you're using native linux file systems with mount points and that doesn't work with NSS--at least it never created any of the stuff there). Is this correct?

2) IF I have failover such that it's possible to have more than one MTA/POA on an OES2 node, I take it I need to ensure that I have the proper .mta and .poa script files on that particular node, right? So I can just run the install again and "configure" for the new agents so that it'll adjust the gwha.conf file properly? Or can I just copy the .poa/.mta files to the /opt/novell/groupwise/agents/share directory?

3) On NetWare, we had to run these in protected memory so that when you migrated/failed a resource it didn't unload all the agents. I take it on OES2 SP2 Linux that there's no need to do this since your load/unload scripts (cluster scripts) specify the exact command such as: /etc/init.d/grpwise stop DOMAIN, etc.?

4) Lastly, on OES2 SP2 linux, if you DO have multiple MTA/POA (diff. resources), do I need to edit the eDir objects OR the .mta/.poa files to specify the actual secondary IP? Or does the load/unload scripts handle that accordingly? We have unique IP and TCP Port assignments for everything, but I just didn't want to get into an odd situation where two unique MTA/POA on the same server started using the physical node IP or each other's IP's.