Re: [exim] Question on IP origination schema

On 16/05/11 13:31, Jeffs wrote:
> Here's the million dollar question: does the final recipient looking at
> the source of the message see that it originated at the IP address for
> Master-Controller.com or the mx mailer for the stated domain?
>
> I need it to be for the IP address of the registered mx domain.
>
> thank you.
>

Each MTA adds a received header, including mater-controller.com's MTA.
This allows for errors to be traced back to their source amongst other
useful things.

Exim has the ability to strip out received headers itself, or you can
pipe the message through some kind of program to change the message
however you wish and re-submit it. You can also set your own format for
received header, but keep in mind that there is an expected format for them.


Why aren't you just submitting the message directly to the appropriate
MTA instead of using a local MTA and then eliminating most traces of its
existence? The software that is generating the messages will have all
the information it needs to submit directly the correct MTA via SMTP
with any credentials it needs.

Of course, even this way, the source IP of the submitting server will be
in the received header added by the SMTP server, unless of course you
have changed the received header format added by that MTA.

Perhaps don't use SMTP for the first step? Any kind of API would suffice
if you want to avoid the origin.

Please note, some people (like myself) use the lack of received path for
emails as a spam indicator (one of many). SpamAssassin gives extra
points to received headers that are unparseable by default too.