RE: Debian Testing Crashes Overnight
Mark L. Kahnt wrote:
>On Thu, 2002-11-21 at 11:26, Michael Kahle wrote:
>> I built the system using Woody, then 'upgraded' to testing on my
>> workstation. I use this as my primary computer here at work. At the end
of
>> the day, when I go home, I lock X and take off. The next morning, the
>> system is frozen. I can't do anything but shut the machine down by
holding
>> the power switch. I tried to ssh into the box but it timed out. So
there
>> was no way to shut it down gracefully.
>>
>> I have XFree86 installed, Gnome installed and am using the mga drivers
from
>> Matrox to support my dual screens. Where do I begin to troubleshoot
this?
>
>You're not running it in a PC emulator on Windows, are you? ;)
nope! :)
System: IBM ZPro, PXeon 933, 768MB, Debian
>
>I've heard of problems with locked screensavers in the past, and I
>wouldn't be surprised if that *could* be a problem, but what is more
>important is whether or not the system hangs at the same time each day
>(potential cron job problem - check the tail of the back logs.)
>Alternately, it could be a hardware problem from poor handling of acpi.
I remember reading something about the screensaver problem to. FYI I am
using the xmatrix xscreensaver. I'm not sure if the system hangs at the
same time each day. If I look in the syslog I see this gap in time occor:
/* Syslog Snip
Nov 21 00:01:31 geocentric dhclient: DHCPREQUEST on eth0 to 192.168.100.254
port 67
Nov 21 00:01:31 geocentric dhclient: DHCPACK from 192.168.100.254
Nov 21 00:01:32 geocentric dhclient: bound to 192.168.100.7 -- renewal in
10800 seconds.
Nov 21 00:08:01 geocentric /USR/SBIN/CRON[2891]: (mail) CMD ( if [ -x
/usr/sbin/exim -a -f /etc/exim/exim.conf ]; then /usr/sbin/exim -q ; fi)
Nov 21 00:21:33 geocentric -- MARK --
Nov 21 00:23:01 geocentric /USR/SBIN/CRON[2909]: (mail) CMD ( if [ -x
/usr/sbin/exim -a -f /etc/exim/exim.conf ]; then /usr/sbin/exim -q ; fi)
Nov 21 00:38:01 geocentric /USR/SBIN/CRON[2920]: (mail) CMD ( if [ -x
/usr/sbin/exim -a -f /etc/exim/exim.conf ]; then /usr/sbin/exim -q ; fi)
Nov 21 00:53:01 geocentric /USR/SBIN/CRON[2928]: (mail) CMD ( if [ -x
/usr/sbin/exim -a -f /etc/exim/exim.conf ]; then /usr/sbin/exim -q ; fi)
Nov 21 09:42:46 geocentric syslogd 1.4.1#10: restart.
Nov 21 09:42:46 geocentric kernel: klogd 1.4.1#10, log source = /proc/kmsg
started.
Nov 21 09:42:46 geocentric kernel: Inspecting /boot/System.map-2.4.18-bf2.4
Nov 21 09:42:47 geocentric kernel: Loaded 19371 symbols from
/boot/System.map-
2.4.18-bf2.4.
Nov 21 09:42:47 geocentric kernel: Symbols match kernel version 2.4.18.
Nov 21 09:42:47 geocentric kernel: Loaded 337 symbols from 12 modules.
Nov 21 09:42:47 geocentric kernel: Linux version 2.4.18-bf2.4 (root@zombie)
end snip */
You see? Nov 21 00:53:01 - cron problem?
Then: Nov 21 09:42:46 - must have been when I got it running again. Because
I had do forcibly shut it down I had errors on my disk. When it crashed I
was doing a dump restore... :( Yea, it's been a shitty morning. So I had
to run e2fsck to check / repair a bunch of inodes. So by the time the
system had actually come back up this morning it may have been that time.
ACPI - Humm... I am loading that module, although I don't have to! BTW
Kernel version 2.4.18-bf2.4, I have not yet compiled a kernel for this
machine.
>
>Could you provide when the last entries in your syslogs have been, and
>anything of interest about your kernel (version, whether you roll your
>own, etc.?)
See above.
Thanks for you time.
Michael
Reply to: