Author: Chris Lightfoot Date: To: Tair Belini CC: exim-users Subject: Re: [exim] Verify problem with forwarded messages
On Fri, Nov 03, 2006 at 04:21:46PM +0200, Tair Belini wrote: > XYZ1 verifies USER1@THIRD1
> since the delivering IP is defined as the SMTP server of domain THIRD1, it accepts message.
> Then, it tries to deliver (forward) message to MAIL2@ABC2
>
> Now, the delivering IP is not the SMTP server of THIRD1, instead it is the SMTP server of XYZ1.
> Thus, server check fails.
I'm a bit confused about all your As and Xs and 1s and 2s
-- if you must obfuscate, can you call them alice and bob
and fish and soup and so forth, so that they're actually
memorable.
Anyway, I suspect that the problem here is that you're
assuming a sender callout verification is made to the
machine that connects to exim to deliver a mail (at least,
that's how I read your last two sentences). This isn't how
it works -- exim tries to route the sender as it would any
other address, and connects to the host that it would
usually send mail to.
(In response to your other comment -- you can forward mail
among different hosts with blank return paths, but this is
only useful in private setups where you know all the
possible drawbacks. Don't do it on the public internet
with unsuspecting people's email addresses. For instance
it will make it impossible for them to distinguish real
mail, real bounces, and bounces resulting from spammers
using their email address as a sender.)
--
``Many people would sooner die than think. In fact, they do.''
(Bertrand Russell)