top and stuck processes

Question

top and stuck processes

Once in a while my server won't wake up the display after sleeping it. I'm left with a black screen and no way to use the system.

If I log in with SSH from another Mac on the LAN, everything seems fine, except that top reports that 1 process is "stuck." But it won't tell me which one! Isn't there some way to identify the stuck process and kill it?

Also, issuing a reboot command remotely doesn't work. The server screen wakes up (displaying the screen saver image) but sits there forever with the little "gear" thingie going round and round.

QuickSilver dual-800 G4, Mac OS X (10.4.8), X Server

Posted on Nov 5, 2006 7:46 AM

Reply

Answer 1

Nov 5, 2006 3:35 PM in response to G Robert Lewis

Hi Robert,
You can use the following command to view all processes:

ps -auxww

In the STAT column is the state of the process. An explanation of the letters in that field can be found in the ps man page. From "man ps":

I Marks a process that is idle (sleeping for longer than about 20 seconds).
R Marks a runnable process.
S Marks a process that is sleeping for less than about 20 seconds.
T Marks a stopped process.
U Marks a process in uninterruptible wait.
Z Marks a dead process (a ``zombie'').

You don't put the whole system to sleep, do you? If you do, it shouldn't even be able to serve anything. Don't put a server to sleep. You can sleep the display but I don't recommend even spinning down the drives.

If you are talking about the display, that's doubtless just a symptom. Without knowing the processes having problems, it isn't possible to know what is wrong. Of course even a listing might not be sufficient to know what's going wrong but it's probably necessary.
--
Gary
~~~~
I'd horsewhip you if I had a horse.
-- Groucho Marx

Reply

Answer 2

Nov 5, 2006 4:22 PM in response to Gary Kerbaugh

Thanks for the info. No, I don't sleep the server or the hard drives; just the display. Interestingly, I found that the problem was much worse when the "Flurry" screen saver module was selected. Right now I'm using "Abstract" and the problem is much rarer.

Quite a few OS versions ago, it seemed I could sometimes recover control by killing (remotely) the Screen Saver process, but that hasn't worked in quite a while now.

Reply

Answer 3

Nov 5, 2006 5:40 PM in response to G Robert Lewis

Hi Robert,
Why don't you skip the screensaver and just put the display to sleep? I imagine that it will forestall the problem but not fix it. Based on a four page thread on the File Service discussion, I gather that AFP servers stop authenticating after a period of time. I actually think the problem is deeper than that and that the AFP server just exacerbates the problem the most. However, that's just a "feeling"; don't quote me on that.
--
Gary
~~~~
To err is human -- to blame it on a computer is even more so.

Reply

Answer 4

Nov 5, 2006 7:01 PM in response to Gary Kerbaugh

I ran the ps command and found two "Zombies". Do you have any insight into what these might be? I'm not knowingly running any python code, and note that the start time is "31Dec69".

root 675 0.0 0.0 0 0 ?? Z 31Dec69 0:00.00 (python)
robbo 684 0.0 0.0 0 0 ?? Z 31Dec69 0:00.00 (python)

Anything to be concerned about?

Reply