How to investigate a system lockup

kvtournh · 12-09-2005, 04:31 AM

Hi,

I don't know if this is the correct forum to put this, so if it is not, I apologize.

I recently had a system lock-up. I could not access over ssh, I could not use keyboard and mouse. So the only thing I could do was a hard reboot.

Checking /var/log/messages and /var/log/syslog I did not see anything suspicious.

Can anyone tell my how to investigate this problem further?

Thanks

klaas

syg00 · 12-09-2005, 05:25 AM

Quote:

Originally Posted by kvtournh

Can anyone tell my how to investigate this problem further?

You can't.
Once you reboot any potentially useful diagnostic information is generally destroyed.

There have been projects to save crash data, but Linux being such an eclectic community, deciding on how/what to save is almost insolvable.
All solutions require some prescience on behalf of the user/administrator in that kernel patches need to be applied in advance of the failure.

The best that can be said at present is that work continues.
There is a way to go before diagnostics will be all encompassing methinks.

nx5000 · 12-09-2005, 05:48 AM

Can reproduce the problem?
Did you change some hardware recently?

Try to go on the console and look at the messages, but the lockup may appear before any message arrives.
Try to test your ram with memtest86

Then the dichotomy logic method (tm):
remove all the hardware you can, try to reproduce.
remove acpi.
shutdown all the software / modules you can, try to reproduce.

I had a lockup yesterday on a server, no mouse, no keyboard, no ethernet led flashing, no message, nothiiing.
Reboot.. Forget..