I'd support work like that, though probably by working on the document more
than the implementation.
I think the name should be OCLIENT in line with DSN's ORCPT, but that's a
minor point.
-MSK
On Sat, Feb 23, 2013 at 6:22 PM, Ned Freed
<ned(_dot_)freed(_at_)mrochek(_dot_)com> wrote:
Hi Ned,
That's only true if you let it be true. There is no reason for the
insertion
of a secondary to interrupt the flow of information to the primary, so
the information available once the message does get to the primary
should
be the same, resulting in the same accept/reject/discard decision being
made.
True, it requires a certain amount of finesse, and some standards in
this area
would definitely help, but there is nothing intrinsicly impossible to
do here.
Actually that is an interesting idea. The backup MX could act as an SMTP
proxy as long as it can maintain a connection with the primary. That
would also sort out your observation on address checking.
And when it can't it needs to be prepared to pass as much of the context
of the
original connection to the primary as possible later when the primary
becomes
available again. The IP address is especially important here, of course.
Again, it would be quite helpful if there was a standard for doing this.
Postfix has XCLIENT, we have something called XPEHLO, and there are
probably others out there.
I would certainly be willing to work on this if there is sufficient
interest.
Ned
_______________________________________________
ietf-smtp mailing list
ietf-smtp(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf-smtp
_______________________________________________
ietf-smtp mailing list
ietf-smtp(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf-smtp