Re: [exim] high (50+) load avg / 0% cpu use / 200+ conn.conn…

Sorry for not mentioning it in my first post... I have default installation
of Debian "exim4-heavy-daemon" package without any filter (spamc/clamd/etc).
The "mad" configs (ramdisk, logging to /dev/null) are all setup as testing
and was made to minimize the possibility of problem's souces (hdd i/o).

Before i launched the system sure i had logs enabled, Maildir located on
separate SATA mounted disk... Nothing interesting were in logs - just ton of
"message received for local delivery", and after some time ton of "refused"
due "too many connections" or "load too high". Only after investigating the
logs i turned them off. Im not that mad and blind, lol :)

Just rechecked exim config... all DNSBLS and other blacklist's are turned
off.

As for now, load average are ok (below 1.0) with about 10k connections per
hour/166 per minute/3 per second, and only <10 exim4 processes. The only
things i changed were "dc_minimaldns" set to true and settings local box's
ip in /etc/resolv.conf (after setting up bind). Default logging is now
enabled.

Will let you know when load avg go high again...

Thanks all for response!

----- Original Message -----
From: "Patrick von der Hagen" <hagen@???>
To: "jamez" <bugtraq@???>
Cc: <exim-users@???>
Sent: Sunday, February 17, 2008 9:02 PM
Subject: Re: [exim] high (50+) load avg / 0% cpu use / 200+ conn.connections


Am Sonntag, den 17.02.2008, 00:56 +0200 schrieb jamez:
[...]
> There're no others services on the box, "top" shows no processes using
> more
> then 0% CPU. us/sy/ni/wa/hi/si/st stats are all low, CPU idle is almost

I think you should have asked earlier. Everything regarding your
system-configuration sounds desparate (not to say: mad) and should be
undone ASAP. Seriously.

Later I'd considere wheter you accidently missed some information and
there are additional processes like spamd, nscd, etc. which might cause
problems when seeing load.


--
CU,
Patrick.