Re: squid 1.1.21 dies with signal 6 every hour

From: Andre Albsmeier <andre.albsmeier@dont-contact.us>
Date: Mon, 30 Mar 1998 07:31:31 +0200 (CEST)

> Andre Albsmeier wrote:
>
> > Here are some extracts of the logfile. Apparently, it gets a SIGSEGV.
> > I think, the hardware is okay since 1.1.20 never showed these problems
> > and they startet from the moment I installed 1.1.21...
> >
> > FATAL: Received Segment Violation...dying.
> > 1998/03/26 14:00:01| storeWriteCleanLog: Starting...
> > 1998/03/26 14:00:02| 4096 lines written so far.
>
> Is there anything interesting before the fatal error in the log file?

Unfortunately no.
>
> To my knowledge Squid does not schedule anything based on hours, and
> your squid seems to die exacly on each hour (or not at all). There are
> some things that are delayed or run with a hour interval, but if this
> was the cause then your Squid should die with a interval of somewhat
> more than one hour, and not on the same time each hour..
>
> Check your crontabs (and cron log) some more. I still think you have
> something running that triggers something in Squid.

No, my crontabs don't do anything on the hour. Also, they weren't changed
and ran perfectly with squid up to 1.1.20.

>
> Since it seems to die exactly on the hour, you could try to enable debug
> logging before it dies (squid -k debug). run squid -k debug at 2
> miniutes before full hour, and we should get a quite clear picture of
> what Squid was doing when it dies (if it does). If it doesn't die you
> can run squid -k debug again to return the logging to normal.

OK, I will try that today.

Thanks, so long...

        -Andre
>
> ---
> Henrik Nordström
> Sparetime Squid Hacker
>
Received on Sun Mar 29 1998 - 21:32:47 MST

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:39:29 MST