spam getting through.

[hermit921]

<x-flowed>
At 09:37 AM 9/15/2004, Ken Goods wrote:
>Michele Neylon :: Blacknight Solutions scribbled on Wednesday, September 15,
>2004 8:36 AM:
>
><*snip*>
>
> > Although I would agree about spamcop I would tend to disagree
> > with the use
> > of SBL+XBL directly in MailScanner.conf
> >
> > I ran a little test on one of our servers the other day for a laugh.
> >
> > The test was totally random and not that scientific, but of the 20 odd
> > emails that had been scored on spamhaus I found that at least 2 of
> > them should not have been blocked outright
> >
>
>Michele,
>You'd be horrified then to learn that I use SBL+XBL in sendmail before it
>even hits MailScanner. It's been this way since May of this year and I've
>not received one complaint. I process 3-8k messages a day for 25 users and
>75%+ of that is spam, lets say an average of 3375 spams a day. SBL+XBL
>through sendmail rejects about 60% of those, or  say an average of 2025
>spams. This is 2025 messages that do not have to be processed by MS, SA, and
>ClamAV greatly reducing the load on the server.
>
>I can only say what works for me, YMMV, but anyone experiencing load
>problems should at least take a look at how SBL+XBL performs in their
>environment and *consider* (carefully) using SBL+XBL at the MTA.

various bits deleted above.

We started using CBL (part of XBL) in postfix several months ago.  We have
three systems, each handling about 35K incoming messages per day.  Last
time I counted, we rejected 35% of incoming connection attempts based on
CBL.  No complaints, either.  I hope to expand to the rest of XBL, but CBL
was the easiest to check at the time.

hermit921

------------------------ MailScanner list ------------------------
To unsubscribe, email jiscmail at jiscmail.ac.uk with the words:
'leave mailscanner' in the body of the email.
Before posting, read the MAQ (http://www.mailscanner.biz/maq/) and
the archives (http://www.jiscmail.ac.uk/lists/mailscanner.html).
</x-flowed>