On Monday 11 October 2004 08:52, Doug Stanley wrote:
> I have a question. Is there some sort of plugin or something that I
> can use like a spamassassin test to check for spf records and then
> adjust the spamassassin score accordingly? Or is this a simple
> allow/deny type thing for now?
SpamAssassin can be useful, but it shouldn't be your first line of
defense against spam. I think the people who came up with SPF
envisioned it being used at the SMTP level. SPF acts on envelope
information, whereas SA acts on the body of a message.
To directly answer your question, I'm sure you can do it in SA (you
might want to ask in an SA forum about it). But envelope issues are
better handled by the MTA itself.
I kill an amazing amount of spam, AFAICT with no false positives[1],
using postfix smtpd_recipient_restrictions (RBL's, access lists, and
simple rules.) Close to 90% of it, I'd guess. This takes place on my
little (32MB RAM) remote virtual server. It uses very little in system
resources, and in bandwidth, for that matter.
> I'd rather start with adjusting spamassassing scores for now than
> completely denying messages right away...
This too can be done at the MTA. Simply adjust the policy daemon (I'm
still speaking in postfix terms) to return a "PREPEND X-SPF-Failed:
blah blah" header for a positive result, rather than REJECT. SA is
definitely not the right tool for this job, even if it can be made to
do it.
[1] I have started using an RBL of dynamic IP ranges, and that will
eventually get a false positive, but the reject message will give the
sender instructions on how to get around the rejection.
--
mail to this address is discarded unless "/dev/rob0"
or "not-spam" is in Subject: header
