I just wanted to let y'all know what the fix finally ended up being.....It
was a couple small things that put together became work stoppers. We were
seeing that the LAN card on the HP was getting resets and we swapped out the
Core I/O card just in case we had a card going bad...problem continued
unabated. Finally, the networking folks admitted that the two major subnets
(204 and 205) were all coming into through one router - and it had an
Ethernet 10mb card. Thats 700 Networked users, 30 networked printers, all
SQL and GQL traffic, all Office product users (networked versus locally
loaded on the PC), and 400 HP users......no wonder we were getting booted!
Our problem stems from explosive growth - the company has grown so fast that
no one ever revisted the original configuration of the network. Next step -
adding a print server.

The fix was to add more subnets, and more networking equipment. The ethernet
cards were upgraded to 100mbit. The Printers were divvied up between
subnets. I want to thank *everyone* who helped - from the HP RC Networking
specialist, John(he was there every step of the way), to Mike Hornsby from
Beechglen who helped with the trace files, and all the responses that I
received from at least a dozen HP listers....y'all are great! We did find
one thing on the HP side that may have contributed to the problem and that
was also pointed out by Mike - apparently the MAU on our DTC was not SQE
heartbeat enabled and we were showing in SOS a system process (I/O Manager)
that was always active and usually our heaviest I/O user. Simply unplugging
the MAU, changing a dipswitch, plugging it back in and doing a download on
the DTC fixed that issue.

I am sure that I oversimplified the Network fix, because I don't think that
the networking folks told me everything. But its fixed, and my users are
happy, my operators have their 15000 page EOP reports printed and my phone
stopped ringing. Thank you everyone!

Lisa Leidig
HP Systems Administrator
Amerigroup Corporation