|
Jared, This pretty much depend on your MTA. For
example, in Exim you can use a System Filter to read the header after SA is
done for the spam score and then do specific actions based on that score. You
could have it delete at one threshold, save a copy at another, and rewrite the
subject at yet another. The possibilities are near endless. Each MTA has its
own filtering abilities. In our setup, we use sa-exim to do all of
this for us at connection time before ever writing it to the hard drive. I hope this helps. Kris From: Jared Priddy
[mailto:[EMAIL PROTECTED] My
company currently receives vast amounts of UCE... SpamAssassin has been doing a
pretty decent job of tagging spam and I am currently working on training the
Bayes filters to make it even more effective. I am not currently not
blocking any mail - for selected users, I have my production mail server forwarding
a duplicate message to my spam filter machine, which in turn runs the tests,
redirects spam to a "junk" mailbox and good mail to a "ham"
mailbox. I currently use these mailboxes to train my Bayes filters,
moving messages to the appropriate mailbox if necessary. When I
begin blocking for real, I will not want keep a duplicate of every mail message
we get but a quarantine will be necessary. I will probably still use a
mailbox for quarantine purposes, but I do not want every message that fails the
tests to be kept. Any message that scores above a certain number, say
20.0, I would like to drop without quarantining since I would be pretty certain
that it wouldn't be spam. Then my quarantine would only contain mail
items that scored between my kill threshold and 20.0. Is there
something built in that enables me to do this, or has anyone monkey'd around
with SA and figured this out? Thanks in
advance |
Title: stop quarantine after sufficient confidence level
- stop quarantine after sufficient confidence level Jared Priddy
- Re: stop quarantine after sufficient confidence lev... Matt Kettler
- Re: stop quarantine after sufficient confidence lev... Cheryl Trooskin
- RE: stop quarantine after sufficient confidence lev... Kang , Joseph S.
- RE: stop quarantine after sufficient confidence lev... Kristopher Austin
- RE: stop quarantine after sufficient confidence lev... Jared Priddy
- Re: stop quarantine after sufficient confidence... 'Cheryl Trooskin'
