Moin,

On Fri, 2025-11-28 at 11:57 +0100, Marco Davids (IETF IMAP) wrote:
> > What did you have in mind to include, Marco?
> Everything that prevents critics from arguing “IPv4 is flawless and
> IPv6 always adds problems” works for me.

Always on board with that. ;-)

> A simple (re)phrasing like this might already help:
> 
> "Broken IP Connectivity at the Resolver
> Similar to authoritative servers, (stub) recursive resolvers may face
> broken IP connectivity, where IPv6 introduces specific challenges,
> e.g., if a client has been assigned..."
> 
> I’ll leave the exact wording to the authors; this is just a
> suggestion.

I took a shot at outlining some points from the discussion in non-
normative language, also highlighting that IPv4 and IPv6 each may face
independent forms of connectivity issues, esp. for resolvers:

https://github.com/ietf-wg-dnsop/draft-ietf-dnsop-3901bis/pull/32/files

Does this work for you? It explicitly picks up Lorenzo's comment about
issues with CGN.

> Regarding my earlier comment on using MUST versus SHOULD in 
> "authoritative DNS servers SHOULD use native IPv6 addresses": I don’t
> want to make a big deal out of it and I understand the rationale for 
> SHOULD. Still, I’d like to note that at the TLD registry I work for, 
> only native IPv6 addresses are accepted in glue records of NS
> servers.
> 
> No Teredo, 6to4, or similar transition mechanisms as we've seen those
> fail too often. We’ve even encountered attempts to use addresses from
> the 64:ff9b::/96 range, which was obviously a bad idea.
> 
> This is why I raised the concern, but I’m fine keeping the wording as
> SHOULD.

Ok, thanks!

With best regards,
Tobias

-- 
Dr.-Ing. Tobias Fiebig
T +31 616 80 98 99
M [email protected]

_______________________________________________
DNSOP mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to