[Mimedefang] SA working but spam still goes through

Kevin A. McGrail kmcgrail at pccc.com
Wed Oct 19 11:42:26 EDT 2005


> here is a header from spam that got through:
> X-Spam-Status:   No, score=0.2 required=5.0 tests=NO_REAL_NAME
autolearn=no
> version=3.0.3
>
> i don't get why it says autolearn=no ??
> when whould it say 'yes'?
see perldoc Mail::SpamAssassin::Conf

and specifically:

       bayes_auto_learn_threshold_nonspam n.nn (default: 0.1)
           The score threshold below which a mail has to score,
           to be fed into SpamAssassin's learning systems auto­
           matically as a non-spam message.

       bayes_auto_learn_threshold_spam n.nn    (default: 12.0)
           The score threshold above which a mail has to score,
           to be fed into SpamAssassin's learning systems auto­
           matically as a spam message.

           Note: SpamAssassin requires at least 3 points from the
           header, and 3 points from the body to auto-learn as
           spam.  Therefore, the minimum working value for this
           option is 6.

> how would i check if a spam email I passed through sa-learn is in its
'db'??

You will see a test that starts with BAYES that will show up in the tests.


And remember, Bayesian Rules are just one test that will help determine if
email is ham or spam.  SpamAssassin has literally hundreds and thousands of
rules available to test with and if you aren't using some of the network
tests look into them.  I'm a big fan of uribl and surbl.

Regards,
KAM




More information about the MIMEDefang mailing list