I sent a message over to the ntpd list about this and they seem to think
it's a kernel issue.
Anyway, so it seemed the lockups of my IPX were showing up when ntpd was
running in high priority mode. Last week I put its priority back to
normal and attempted to break it again by polling it every five seconds.
Today I got my result, ntpd itself locked up but according to ps, its
status was "uninterruptible sleep". This time, because the priority was
normal, the entire machine didn't crash but the cpu was spinning wildly
until I was able to kill and restart ntpd. I forgot to try and run a
trace on it, but I'm sure it'll lock up again (ktrace -p <pid> does it
correct)?
Any ideas why this would happen? I'm assuming that the lower priority
prevented the locked up code from taking over the entire system.
Everything else stayed running but ran very slowly (five seconds from
keypress to character on screen) until ntpd was restarted.