[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: crash analysis?



" liStan Hoeppner wrote:
Miles Fidelman put forth on 6/10/2010 6:27 AM:
Hi Folks,

My system is doing something funny, where it seems to be crashing once a
day, and there's not enough in my logs to figure out what's going on
when it crashes.

It looks like most of what I find about crash logging tools (by
googling) is about 5 years old, so..... can anybody offer more recent
suggestions as to logging settings to change, tools to install, and so
forth for capturing and analyzing crash events under Debian (running
Lenny).
Desktop or laptop?  Make and model please.  Is this PC in a "hot"
environment--is it over heating?

What application(s) are you running when it crashes?

Define "crash".  Does it hard lock/freeze requiring a power cycle?  Does it
auto reboot?  Other?
Server, custom configuration from Rackmount, Pentium chip, somewhat old. In a datacenter with conditioned air.

Mostly it's running a mail server, list server, database and web server, but... I recently virtualize the environment and now I'm seeing these periodic crashes. Running Xen, with a couple of DomUs, one containing the old environment.

I suspect an out-of-resource panic, possibly caused some cron job or other, but I don't currently have visibility into what's going on just before one of these crashes - I can't find anything useful in the log files to suggest what's happening when it dies - one moment it's chugging along, the next there's a several second gap followed by a line containing “log source = /proc/kmsg”

Hence, my interest in installing something like LKCD or kdump to capture more data at the time of crash/reboot, or maybe something to capture a rolling log of what's running - anything that can give me insight into what's happening just prior to one of these crashes.

Which brings me back to my original observation that all the documentation, how-tos, and what not, related to crash analysis, seem to be 5 years old or more; and my question as to what might be best current tools and practice.

Thanks,

Miles Fidelman


--
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra



Reply to: