I have a netware 6.5SP8 server that starts to die 6 hours after being
started. Anyone have any ideas as to what process starts exactly six
hours after the server is started?

This server has been running like a great Netware server should. It's
extremely reliable, and never really had any problems. It's on a HP
DL380 rack box. About 10 days ago there was a major abend that
caused a lot of damage. The origional abend was in java.nlm This
abend trashed DS on the box where 90% or more of the objects were
gone, and it was in a tree of it's own. Ended up having to remove DS,
put it back on, and then fix everything because of that.
(certificates, home directories....) After that, everything is
working, seems to be performing well, except we have what I will call
a 6 hour problem. Since this problem, I've reinstalled SP8 plus the
extra's patches, swapped all the hardware with an identical system,
except the drives. Also reran the HP Proliant Support pack. Server
was fully updated about three or four months ago, and has run well. A
patch or two was added for iPrint since then.
HP tools show everything is good. NSS Poolrebuild shows all is well.
Server works without issue for exactly six hours. Late last night
when no one was on the server, I logged on. The server utilization
was from 1-3%. I had a continous ping to the server from the
workstation, and all packets were getting responses. Right at the 6
hour mark, the server went up to 6% utilization and sat their for
about 10 seconds, then back down again to 1-3%. At that time, a few
pings were now timing out. As time went on, more and more pings
failed. After about 30 minutes, there is no longer any real
connectivity with the server. Probably one out of fifteen pings
received responses. Utilization stays in the 1-3% average, no errors
on the logger, NRM shows all is fine, HP Insight Management shows all
is fine, ilo shows all is fine. The network adapter shows traffic
being sent and received. ECB's are ok. I'd say 70% of the time, I
can go to the server and reboot it without forcing it. However, it
does take it a long time to go down. (The other 30% of the time, I
give up and force it down. If I'd let it go for an hour, it may
shutdown on it's own.) When going down, utilization stays low, again
average 1-3%. Changed out the entire box, except the disks. So the
network card isn't at issue. Swapped the network cable, and ports on
the switch with another server and no change. Network switch shows
connectivity good, no problems / errors. Something runs at the six
hour mark, and maybe if I can figure out what that it, it may point me
where to look to find what is at issue.

Any thoughts? Thanks!