Re: SMTP and IPv6 (was: Re: IETF mail service outage planned for 120 0 UTC on 27 June 2024))

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi John,

I should have made it clearer that I was referring only to the difficulty at the integration level, not the technical level (which is completely above my competence level).
My point was that supporting ipv6 for e.g. a website / API is very easy as 1) it's almost universally supported by all cloud providers and 2) even if that wasn't the case, adding a proxy somewhere in the middle is trivial.

I doubt the same can be said about outbound SMTP. The entity sending the emails has to support ipv6 natively. It's not something you can just patch in with AWS SES or other services.
My understanding is that it's an all or nothing situation. Hence the need to find a provider that supports sending emails over ipv6 out of the box.

Nick


On Sat, Jun 29, 2024 at 12:00 AM John C Klensin <john-ietf@xxxxxxx> wrote:


--On Friday, June 28, 2024 19:52 -0400 Nicolas Giard
<nick@xxxxxxxxxxxxxx> wrote:

> For context, absolutely none of the major email delivery providers
> (AWS SES, SendGrid, Mailgun, Postmark, SMTP_com, etc.) support ipv6.
>...

> The email world is unfortunately still overwhelmingly ipv4 and
> unlike other protocols, adding ipv6 support is not trivial.
>...

Nicolas,

Because of a probably relevant document I'm working on in/for the
IETF [1] and my roles with several of its predecessors, the portion
of the above statement starting with "unlike other protocols..."
strikes me as a bit strange. SMTP started out as a protocol running
over NCP and was migrated to TCP / IPv4 with relatively little pain.
That implies some cleanliness of design.  My understanding has been
that several SMTP MTAs have been adjusted to support IPv6 and that
the most important problem since mechanisms for use with IPv6 were
specified in RFC 2821 (in April 2001) has involved keeping up with
subtle changes in the preferred presentation form of IPv6 address
literals.

The mail format specification (RFC 5322) is even less sensitive to
details of the underlying transport.

So, could you explain what you believe makes it more difficult to add
IPv6 support to email than "other protocols"?

Now, if instead of saying "email" you has said something about
two-way IPv4 <-> IPv6 email gateways or issues with mail submission
systems or MUAs that are implemented in ways that are inconsistent
with the spirit and intent of the standards, that might be a
different story entirely.  But, if you see problems with the core
email protocols and IPv6, I, and presumably the EMAILCORE WG would
like to know about them... and soon.

thanks,
   john

[1] draft-ietf-emailcore-rfc5321bis


[Index of Archives]     [IETF Annoucements]     [IETF]     [IP Storage]     [Yosemite News]     [Linux SCTP]     [Linux Newbies]     [Mhonarc]     [Fedora Users]

  Powered by Linux