Four node cluster. NW6 sp4. DELL Servers, EMC CX400 SAN.

Six resources: Mail1 & Mail2 (GroupWise), Vol1, Vol2, Vol3. Vol4

One of the volumes started developing corruption. Every night, one of the
servers would abend NSS Free Tree(3) is corrupt, and say ZLSS.NSS caused the
Abend. The problem got worse it seems when we ran a purge on what we
identified as the volume with the problem. The 1st node tried to bring the
volume online, and abended - and went down the line abended all 4 nodes.
NSS Rebuild did fix the volume at least temporarily. We've got some
upgrades to do to get to NW65, then a DELL SAN Health Check Upgrade SAN
software and our drivers - are planned.

In the interim, we learned some lessons, and thought it might be wise to ask
about our plans to help prevent this from happening again.
1. We've made certain resources restricted by node. ie Email can be on Nodes
2 and 4.
2. We are thinking about the following plan - and this is my question -
Move to having 2 clusters instead of one.
Cluster 1
Vol1 - restricted to CS1 + CS2
Vol2 - restricted to CS2 + CS3
Vol3 - restricted to CS3 + CS4
Vol4 - restricted to CS4 + CS1
Cluster 2
Email1 - restricted to CS5 + CS6
Email2 - restricted to CS6 + CS7

Besides for the obvious cost in licensing and hardware, any thoughts?