Just looking at the OOPS message makes me suspect I hit a kernel bug.
I ran a memory tester when I originally acquired the machine. It's
possible that some flaw has developed between now and then or we got
hit by a particularly bad cosmic ray.
It's funny, hardware generations change so quickly. I
BTW do you have ECC-enabled memory on that box? If not, it will probably
makes sense to run memtest86 (from www.memtest86.com) for at least few
hours on that box (hmm, this will at least be *planned* downtime :).
My own linux kernel testing experience tells that the first thing to do
after kern
Mail-Archive went down this morning arond 6am Pacific time and was
offline for several hours. When I eventually noticed, I went over to
the facility to poke around. The computer had crashed, with a nice
"OOPS" message from the linux kernel that looked to me like a null
pointer in the virtual memo
