Have you tried logging onto the system from another machine via ssh or telnet during these periods?
If so, is it sticky from there too? I'm wondering if something could be hanging up your X server.
Also, are there any hints in /var/log/messages* or /var/log/boot.log*? Nothing jumps out at me in
your top log. You might try logging "ps -eaf" and "free" to get an idea about
memory usage and all processes listed on the system.
On Wed, 11 May 2005, Gabe Langbauer wrote:
So just for fun yesterday I disabled the cleanup script, just to see what
would happen. What happened was that the computer froze like always, only
this time I didn't have the runaway "gf" process.
Locally the computer was responsive enough that I was able to get it to
reboot cleanly, however unresponsive enough that even given 2 full hours
it was unable to open a terminal
When I rebooted I got an error with rtstats, but wasn't quick enough to
actually get the exact error.
I logged the top command, and it shows a few things:
1) The logfile continued well after twister became unresponsive
twister stopped responding around 4:45, the logfile continues until 7:45
(when I rebooted)
2) Nothing is very taxing to the system
3) The only running process is the ldm
Upon reboot the ldm product queue was corrupt.
I've attached the top logfile in case someone has a better idea how to
interpret then I do