Dear Robert et al,
Robert Brooks wrote:
[..]
> Mark,
>
> many thanks for this. I'd not thought about the
problems with
> performance. However it is precisely because this is
prequeue that I am
> interested.
Me too, I'm interested to have some of the tests done
pre-queue.
>
> I'm keen to avoid accepting email which is not going to
be delivered.
> I'd rather the server which has the message is
responsible for bouncing
> than us creating back-scatter or silently failing to
deliver.
Thats one point for which I'm interested in pre-queue
filtering
>
> Also ultimately I'd like users to be able to chose who
they do and don't
> accept email from. Just because something is
blacklisted doesn't
> necessarily mean it's spam (or that all users wish this
address to be
> blacklisted). There can be personal reasons for
blacklisting or
> commercial emailers that aren't diligent with
unsubscription requests.
I turned off BL filtering in SA and enabled them in Postfix
(reject_rbl_client bl.spamcop.net etc.). This checks are
relatively
cheap, it does not questioning the expensive SA.
I also applied some basic sender and header checking with
postfix which
are also cheap and quite successful.
The result of this is quite a lower load on the mailserver
and properly
rejected mails for the methods which are
"dangerous" for false positives.
Hope this helps.
Thanks,
Luc
------------------------------------------------------------
-------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the
chance to share your
opinions on IT & business topics through brief surveys -
and earn cash
http://www.techsay.com/default.
php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
AMaViS-user mailing list
AMaViS-user lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/amavis-user
AMaViS-FAQ:http://www.amav
is.org/amavis-faq.php3
AMaViS-HowTos:http://www.amavis.org/ho
wto/
|