We installed 4 new servers running Netware 6.5 SP1.We tested AFP access
from Mac OSX 10.2.8 and found it didn't work at all (we already had it
working fine on our Netware 6.0 SP3 servers).So we patched the 6.5
servers to AFP Version 2.01.12 17 March 2004 (the latest Beta
patch).This appeared to fix the problem.

BUT we have noticed that one of our new servers regularly has a much
higher than normal CPU utilisation state caused by what appears to be
several very busy 'stuck' AFP threads.This effectively prevents Mac OSX
from logging into this server.This was never a problem in Netware 6.0
SP3.The only solution for this is to force the server down (ie power off
as it won't go down cleanly ) and restart.

1) Is this a known problem with the Beta patch for Netware 6.5?
2) Is there another patch coming out to resolve this issue?
3) Can we safely backrev to a working Netware 6 NLM? (not the best
solution as we would lose the obvious AFP v 3 benefits).
4) Would upgrading the TCP stack work?