SCO OpenServer 5.0.5 Problem

I am currently having a problem with my SCO OpenServer 5.0.5 / 3.2v5.0.5 system.

Running smoothly for the past 3 years, at random on March 16, 2010 the system crashes. I reboot it. Everything is normal again. 9 hours later it crashes again. This cycle has repeated itself ever since then. System is crashing every 9 hours and I have to reboot it.

This is the following error message that I get on the screen after it crashes:

Unexpected trap in kernel mode
Panic: k trap - kernel mode trap type0x000000E
tying to dump 32671 pages hd (1/41)
409 pages per '.'

This system is critical to my business and I cannot have it crashing all the time.

This is what I have done so far.

  1. Replaced Motherboard
  2. Replaced CPU
  3. Replaced Power Supply
  4. Replaced Ram
  5. Replaced Network Card
  6. Rebooted all network equipment (3 Switches and 1 Cisco Pix Firewall)
  7. Replaced hard drive and copied old data from a tape from the night before (3/15)

The strange thing is, which makes me think it is a network or software issue, is that when I disconnect the system from the network and let it run, it does not crash.

I cannot see or think of any devices on my network that would randomly every 9 hours cause my system to crash when it was running 3 years no problem.

I have exhausted all of my resources, I refuse to pay $600 to SCO for a help session that may or may not fix my problem.

If anyone could help me, I would greatly appreciate it.

Can you post the actual hardware?
Is the interval between crashes exactly the same?
Do you have cron jobs that coincide with the crash?
Does the dump complete, or do you run out of swap space first?

And the results of "df -kv" and "crontab -l" (L).