Arkadiusz Miskiewicz wrote:
> On Sunday 17 of May 2009, W B Hacker wrote:
>
>> Queue runners will often collide when run much under 45 seconds. Ordinarily
>> harmless, save for more log lines than useful.
>
> Collision is not a problem.
>
>> ... then multitail / wtail the exim main and panic logs,
>>
>> plus /var/log/all.log /var/log/messages
>>
>> (watching for abnormal termination of exim offspring).
>
> I just looked over exim source and SIGSEGV for example isn't catched and
> logged :-( Same for other signals.
>
> Would be cool to catch these nasty signals and log these + log backtrace() for
> SIGSEGV.
>
>> There may *already* be clues in those other logs if not in Exim's, so once
>> you have a pattern you can grep out the history and timing.
>
> There are none unfortunately.
>
>> Bill
>
>
Welll .... my Ecim(s) ordinarily stay up until I intentionally stop 'em, as in:
grep -r SIG /var/log
/var/log/exim/mainlog:2009-03-19 22:50:10 [5006] SIGINT received while reading
local message
And/or the OS might be able to log them even if Exim is not...
Do you have (or the Linux equivalent of..) these?
(from *BSD /etc/syslog.conf for an unattended or 'headless' rackmount)
=====
*.err;kern.warning;auth.notice;mail.crit /var/log/console.log
.
.
.
# uncomment this to log all writes to /dev/console to /var/log/console.log
console.info /var/log/console.log
.
# uncomment this to enable logging of all log messages to /var/log/all.log
# touch /var/log/all.log and chmod it to mode 600 before it will work
*.* /var/log/all.log
====
Bill
