Best way to use blacklists?

Timothy M. Lyons lyons at digitalvoodoo.org
Mon Nov 18 20:48:05 GMT 2002


I personally use multiple blacklists - however I only use one of them
with sendmail (relays.ordb.org), I then implement spamcop.net and
Infinite-Monkeys within MailScanner.

I find that ordb is quick and gives me a pretty good coverage, plus I
have had very few problems with false positives - those that I have
noticed, have been resubmitted and cleaned out quickly.  Spamcop however
tends to be a bit too aggressive and has inadvertently blocked out
numerous valid senders so I wont use it on my server other than with
MailScanner/SA to indicate the possibility.

Just my .02

--Tim

-----Original Message-----
From: MailScanner mailing list [mailto:MAILSCANNER at JISCMAIL.AC.UK] On
Behalf Of Richard Siddall
Sent: Monday, November 18, 2002 15:39
To: MAILSCANNER at JISCMAIL.AC.UK
Subject: Re: Best way to use blacklists?


Bill Anderson wrote:
>
> I notice there are several ways to utilize blacklists.  Looks like you

> can do it in mailscanner, sendmail, and the other option is to let
> Spamassassin do it.  Any ideas on the best way to implement with pros
> and cons would be greatly appreciated?  Thanks.

!!!FAQ Alert!!!

I guess my view is that it depends on your confidence in the blacklist:

1/ If you absolutely trust the blacklist and want to reject the mail,
false positives and all, use it in sendmail.  That will result in the
least processing load too.

2/ If you absolutely trust the blacklist, but want only to flag the
message as spam, use it in MailScanner.

3/ If you have mixed feelings about the blacklist, but consider it
useful in detecting spam, use it in SpamAssassin (and consider tuning
the score SA assigns).

I hope this helps.  Bear in mind that the more blacklists you use, the
longer it takes to process each message.  If you're rejecting mail in
sendmail, the sender can time out while waiting for you to check all the
blacklists.

Regards,

        Richard Siddall.



More information about the MailScanner mailing list