Hello myname,

Friday, August 22, 2003, 10:33:44 PM, you wrote:


m>   Since I control the server where the mails are intitally received I
m> can just block these spammers at the gate.  I will get a list of all
m> domains in the mail from , and create an Obviously spam domain list and
m> then block these domains, If I am getting too many mails from these
m> servers.  This way I avoid all the trouble of receiving the mail on my
m> local server and then scanning them

No, that won't work because spammers typically spoof the
"from" field - so you will end up blocking the innocent
victims of the spammers. Since SpamAssassin also effectively
stops many viruses, such as Sobig, if you don't have a
separate virus scan before SpamAssassin, you will end up
with all the stolen email addresses worms like Sobig & Klez
use, too.

If you want to do what you propose, you will need a program
that verifies the spammer domain against the IP, via a
reverse DNS check, and only blocks those that actually
emanate from the domain.

But I have to tell you, it will be a neverending task,
because the spammers keep coming up with new domains. So it
is doable - once you identify a persistent spammer, you can
use the Sendmail access file to block them - but it will
still affect a relatively small number of spammers. I have
almost 1,000 lines with domains that are blocked via my
access file, and out of more than 2000 lines in my log
sendmail log file for the past 20 hours, only 21 emails were
blocked that way - and I would guess that no more than 7
different spam domains were represented in that. That
compares to 527 emails that were blocked because of open
relay RBL listings, 139 that were identified as spam by
Spamassassin, 84 that were cleared as non-spam by
Spamassassin, plus probably at least 200-300 identified as
spam or viruses by other filters on my system.

I know it feels good to block some of these domains - that's
why I do it -- but I figured out a long time ago that I was
just working very hard to duplicate the information already
compiled by various RBL lists. Basically, listing bad
domains on a block list is the LEAST EFFECTIVE approach for
me. However, I have had very good luck with a procmail
recipe running ahead of Spamassassin that filters out all
email with what I consider to be a "spammy" word in the
domain name or from field -- words such as "offers" or
"optin" or "deals". Those are wonderful because they catch
the new domains the spammers come up with - (example from
today: "smartlowmortgagecenter.com "). But you have to
filter on those rather than block, because there is more of
a chance of a false positive.

You can, of course, also create a Spamassassin recipe to add
points for the same thing, I just happen to use Spamassassin
on the tail end of a lot of other filtering routines.

-Abigail



-------------------------------------------------------
This SF.net email is sponsored by: VM Ware
With VMware you can run multiple operating systems on a single machine.
WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines
at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0
_______________________________________________
Spamassassin-talk mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/spamassassin-talk

Reply via email to