On Wed, Feb 04, 2004 at 05:22:20PM +0100, Daniel Tiefnig said:
> Patrick von der Hagen wrote:
> >I agree, clamd is a stability-nightmare. On my system I use "monit"
> >to detect failures and restart clamd as needed. However, I have a
> >much lower volume, so clamd crashes only once a day. On your system
> >that probably wouldn't be a solution.
>
> Ah, yes, we had a script here, that did a "PING / PONG" test every
> minute, resulting in clamd being restarted nearly every minute or so.
>
> >I'm a littel bit surprised no one else here mentions such problems?
> >Or do they have solutions I'm not aware of?
>
> Maybe just not much load on the virusscanners..
It helps considerably if you run the mime acl before clamav, so clam
only has to scan the zipped parts, and things like .exe and .pif are
rejected before they ever get to clam.
Also, people may be interested in knowing that I'm currently packaging
up new clam packages for debian, that include a lot of cvs code to fix
stability issues - they work well for me, but YMMV. Available right now
at
http://www.lobefin.net/~steve/debian.html - the packages are under
the sid/ tree, but build well on woody as well.
These are preliminary packages, and do have some bugs (especially in the
milter package), so caveat downloader. Any patches or suggestions
welcome, if you want to give them a try. Please be patient with
downloads - it's not that fast a link :)
--
--------------------------------------------------------------------------
| Stephen Gran | We're all in this alone. -- Lily |
| steve@??? | Tomlin |
| http://www.lobefin.net/~steve | |
--------------------------------------------------------------------------
