We have a 4 node NW 6.5 NCS cluster that we need to change the IP address
on. The nodes also hold eDir replicas but there are other servers that
also house replicas of the same partitions. Normally, I try to only
change one server IP at a time to allow the limber process to discover the
change. Since NCS requires all nodes to be on the same subnet I see no
way to accomplish this IP change without an outage. Here's my plan:

1. Shut down NCS on all nodes...bring down the cluster.
2. REM out "LDNCS" in AUTOEXEC.NCF.
3. Change IP address on one server at a time and reboot.
4. Make sure DS is properly communicating on new IP address on all nodes.
5. Fix related IP addresses on each node using NRM.
6. Restart NCS on each node. Make sure cluster comes up normally.
7. UnREM "LDNCS" in AUTOEXEC.NCF.

Does this sound like a reasonable plan? I'm open to suggestions if there
is a better way. Thanks,

-Chuck, CNE6