Among many other problems with our production SCO OpenServer 5.0.5 system, it kernel traps often. Those who set it up did not allow enough swap/dump space to capture an entire dump - only 16000 pages of 260000. Whenever the machine panics, I take note of all the register values; I've got over a dozen of these lists just since last April. Is there anything I can tell from this data without the dump?
There is some information on this in the OpenServer Handbook re: comparing EIP numbers to determine software/hardware cause, error messages in /usr/adm/messages, etc., but I'm looking for more detailed information.
And yes - on the new 5.0.7 server we're working on, there's more than enough swap/dump space...
I can tell at least one thing: the panics this morning and three days ago were caused by the same condition, I believe: the register values are identical. Don't think I've ever seen that before.