> > The machine seems to last only about 15-30 minutes between reboot and
> > lockup. Clearly an unworkable situation. Here's a machine
> > description:
> > ALR Revolution 6x6 with 4 Pentium Pro processors installed
> > 512 MB RAM
> > 3COM Etherlink 3c905 network card
> > Fully static kernel (no modules or kerneld)
> Rough guess:
> Are all the PPro's the same stepping?
Yes.
> Are you sure the memory you are using is good (looks like it might be memory
> corruption)?
Memory passes power-on test, and everything works fine with only one
processor enabled.....
> Does it do the same type of behavior under another OS?
Never tried another OS! (You mean there are other OSes? :-) )
> Have you tried a recent 2.1 kernel (which has cleaned up SMP)?
AHA! This is what worked! Based on this suggestion I got the latest
kernel: 2.1.120. No more crashes! Unfortunately, for some reason
networking stopped working (couldn't set up the route tables). Based
this and some e-mail I've seen here, I'm not overly confident in
2.1.120, so I backed it off to 2.1.117 (someone had mentioned they used
that one and it was stable), and everything seems to be working ok
now. It's been up for about an hour now with a constant load average
over 2, and not even a hiccup. So it looks promising!
Anyway, I'm not sure what was wrong with the other kernels, and if
anyone here wants to try to figure out what went wrong. I can send
more information if someone wants to try to figure out what was wrong
with 2.0.35, but for now it looks like I'm pretty happy.
> If all else fails, you could always just send the box to me as an educational
> donation.. ;-)
Sorry -- it looks like it's actually working now. That's what you get
for being helpful! :-)
--Steve Tate srt@cs.unt.edu
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/faq.html