From: Adam (suckmy-balls.com)
Date: Mon Nov 01 2004 - 21:56:46 CST

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On Mon, Nov 01, 2004 at 09:22:30PM -0600, Kevin wrote:
> On Tue, 02 Nov 2004 10:53:32 +0800, Lars Hansson <larsunet.net.ph> wrote:
> > Adam wrote:
> > > This would definately lead me to believe you don't know what high volume
> > > mail is. ClamAV is also not stable on OpenBSD under heavy mail loads,
>
> Please tell us what you mean by "high volume" mail. 20K? 200K? 2M?

Per what? I don't know what it did per day, probably not too much. It
was 20k per hour during peak times though. And every couple of days
it would segfault, always during peak hours like first thing in the
morning when everyone was replying to all their mail.

> What are you using to interface SMTP to clamd? milter?

Milter.

> > Maybe you define heavy differently but I've put 20k+ mails/day
> > thru clamd for more than a year and it has not once caused any problems.
> > Of course, I run it supervised so it might have crashed now and then
> > but it has not *ever* hung.
>
> Same experience here.
> I run 'clamd' supervised and systrace'd, and while clamd consumes
> quite a bit of memory, I have yet to see it hang. Usually 'clamd'
> runs until I upgrade ClamAV (the engine, not the signatures, freshclam
> gets those once a day)

Clamd itself, and the milter would both occasionally crash. I am not
the only person to have seen this, in fact if you search the archives,
you should find a patch someone worked on to clean up some of the mess
in the clamav code to make it more stable. It may or may not have been
fixed since I last used it however, as they did seem to release new
version fairly often. Last I used was 0.74, and all the 0.7x versions
had this problem, much worse in the earlier versions.

Adam

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]