We have a pair of servers running netware 6.5 with the latest service
pack installed 4a, these servers are clusterd with both connected to one
large raid array, the problem that we have is that the server that is
hosting the volumes will gradually increase in CPU utilization eventually
sticking at around 40% at this point users can no longer access the
volume, the way to rectify this is to force this server to leave the
cluster so that the volumes are then switched across to the other server
and then restart the server that is stuck at 40% utilization.Unfortunatly
this pattern is then repeated on the new server that is hosting the
volumes after a period of time.
Has anybody any idea what could cause this to happen there are plenty of
tids on high utilization but nothing on this particular problem.

Any help appreciated