Re: [exim] Anti-SPAM detection LAB

On 06/01/2008 14:20, tall27 wrote:
> Hello,
> I try to use Exim as a hub SMTP server, that should forward all incoming
> e-mails (for a list of predefined domains) to a number of back end servers.
> The back end servers runs different Anti Spam products - and my goal is to
> check those different products. For this I would like that Exim send
> messages as much as it's possible with the same format as it was received by
> Exim ( without add any new lines and perform changes to the e-mail) - in
> other words - internal servers should "think" that they are receive e-mail
> from the Internet.
>
> What could be the most effective config to archive my goal?

It's going to be tricky because many of the valuable things you can do
to test for spamminess you can only usefully do if you're the edge
server (where your exim is in this case), e.g. reverse DNS lookups,
delays/tarpitting, checking proper SMTP compliance, etc.

On the other hand if you're just comparing content scanners, exim
doesn't change message contents, and would usually only add a received:
header, which you could turn off (by unsetting received_header_text) but
even then a good content scanner will notice the missing header from the
intermediate server (your exim) and score against it.

Overall I think you need to bite the bullet and put your test
configurations right on the internet, not behind exim. And set up an
exim with an anti-spam configuration: you'll probably find a well-tuned
exim+clamav+spamassassin beats all the other anti-spam offerings :-)

Cheers,

John.