> Has anyone experienced hard lockups under stress lately? It allways 
> happened when the machine was on for some time and it was doing the nightly 
> audits. The version that would lock up now and then - with no coredums :( - 
> was about 1.6C. The newer 1.6ZC i run now works fine allthough the 1.6 X 
> server dies regulary. I haven't looked into detail yet but it might be 
> COMPAT_16 now i think of it... i'll have a look in GENERIC.
> 

I was getting repeated lock-ups when I had an Ether3 card fitted 
(averaging about 1 a day).  I switched back to an Ether1 on Saturday and 
the machine hasn't crashed since, despite doing operations which seemed 
guaranteed to lock the card up.

The problem is that although I think I could repeat the problem on demand, 
there's not much I can do to track it down since I can't get at the 
console when X is running (might be able to get serial console working if 
it would really help, but even that won't be too much help since the other 
machines lock up on NFS as soon as my RPC freezes).

> Could any of you also check if the floppy interface works? Mine is prolly
> broken so any responses either if it works or not would be most apreciated.
> 

It didn't when I tried it a couple of weeks ago, just froze the machine 
entirely.  Haven't tried it since switching Ethernet cards.

R.