Craig Morrison wrote:
Gary V wrote:
Exactly. How you prevent sending the message through SA is not a
function of SA itself, but of the implementation, and because of the
large number of implementations and configurations I question whether
it would be practical (or even related) to provide examples of the
various procedures.
Point well taken Gary.
I didn't see much of anything on this subject in the Wiki.
Neither did I.
I've been googling a bit and the cornucopia of hits for
<insert_your_MTA_here>+spamassassin is a mess. :-)
My solution to this problem is this:
I'm running postfix 2.1.5-5 on Fedora Core 3 and recently had this same
question come up. I was whitelisting all 30something domains I hosted
but ran into spammers using foo@<domain name> to get around spam filtering.
My solution was to create a rule in postfix main.cf:
smtpd_recipient_restrictions =
permit_sasl_authenticated,
check_client_access pcre:/etc/postfix/non-auth.re,
-- snip --
And the contents of non-auth.re is:
/^/ PREPEND X-No-Auth: Unauthenticated Sender
Since postfix aborts checking at the first match, this has the effect of
stamping every single message with a header that I can find and react
to, which for me bypasses spamd -- note: this decision takes place
_after_ virus scanning, etc.
Hope this helps someone.
-=Ray
