I have a customer with a 3-node NetWare 6.5 (SP3) cluster presenting fileservices to an 18 server Citrix farm.

This Cluster functions correct, but in event of a fail-over or migrate all of the Citrix servers are going to 100% CPU utilization and have to be rebooted to function again.

Is this normal with Citrix?

Can I do some tuning to eliminate this problem?