On Tue, Mar 07, 2017 at 03:15:10AM +0000, Viktor Dukhovni wrote:
> On Tue, Feb 07, 2017 at 05:27:48PM +0000, Salz, Rich wrote:
>
> > I put the time period as six weeks, which takes us to just around IETF-98...
> >
> > PLEASE reply on list if you will review and/or are interested in working on
> > interop.
>
> I see there's no reference to use of DNSSEC resolvers by CAs that
> implement DNS challenges. Just a suggestion to send probes from
> multiple networks to avoid MiTM attacks, which seems rather weak
> to me. The MiTM might be collocated near the victim rather than
> the CA.
>
> There was some brief discussion of DNSSEC back in Oct/2015:
>
> https://www.ietf.org/mail-archive/web/acme/current/thrd3.html#00561
>
> https://www.ietf.org/mail-archive/web/acme/current/msg00561.html
> https://www.ietf.org/mail-archive/web/acme/current/msg00562.html
> https://www.ietf.org/mail-archive/web/acme/current/msg00563.html
> https://www.ietf.org/mail-archive/web/acme/current/msg00564.html
> https://www.ietf.org/mail-archive/web/acme/current/msg00565.html
> https://www.ietf.org/mail-archive/web/acme/current/msg00729.html
>
> but no further action. Is there a compellng reason to avoid
> requiring acme CAs to spin up a validating resolver? It does not
> seem like a lot to ask. If a domain is DNSSEC-signed then its ACME
> challenge should IMHO be validated via DNSSEC.
Specifically, it 10.3 use of DNSSEC is a RECOMMENDATION, not a
requirement:
https://tools.ietf.org/html/draft-ietf-acme-acme-05#section-10.3
I would have expected a requirement here.
--
Viktor.
_______________________________________________
Acme mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/acme
