[dns-operations] DNSSEC deployment incentives

[Viktor Dukhovni]

On Wed, Jun 19, 2019 at 08:27:34PM +0200, Tom Ivar Helbekkmo via dns-operations wrote:

> Paul Vixie <paul at redbarn.org> writes:
> 
> >> On Tue, Jun 18, 2019 at 9:33 PM Viktor Dukhovni <ietf-dane at dukhovni.org>
> > ...
> >> > But Let's Encrypt de-monetized certificate issuance, so now that
> >> > obstacle is moot.
> >> 
> >> It has also eliminated the incentive to deploy DANE for free certs.
> >
> > not for me, but i think you may be right in general.
> 
> OK, I'll bite.  My impression has been that DANE is unwanted by the
> large makers of browsers because their owners also earn money from the
> CA business, and widespread use and acceptance of DANE would teach their
> customers that they don't necessarily have to pay for certificates to
> achieve what they want.
> 
> So why would the presence of Let's Encrypt lead to *less* use of DANE?

I don't see any evidence that DANE users were doing it to save
money, even before Let's Encrypt came online.  I would expect DANE
adoption to be driven by an interest in securing the underlying
service.  A large fraction of DANE-enabled MTAs (like yours) have
Let's Encrypt issued certificates, this is even true for one.com
whose DANE-enabled MTAs serve ~700k customer domains, and use
Let's Encrypt certs (ACME simplifies cert rollover) with "3 1 1"
records.

The frequent certificate rollover with Let's Encrypt does create
some issues for the sloppier SOHO DANE implementers, because some
forget to implement a robust process for keeping the TLSA records
in sync with the frequently updated certificate chain.

One thing I'd like to see is the ability for a server to use its
current private key and certs (that match its existing TLSA RRset)
to use DNS "update" to publish a revised TLSA RRset (perhaps via
SIG(0) with the TLSA records used as the ACL in the nameserver).
If any of the implementors of BIND, NSD, PowerDNS, ... are interested,
please reach out...

Better tooling would help, I believe EFF have a work-in-progress
to make certbot play nice with TLSA records.

    https://mail.sys4.de/pipermail/dane-users/2018-February/000440.html
    https://github.com/danefail/list/issues/47#issuecomment-456623996
    https://community.letsencrypt.org/t/please-avoid-3-0-1-and-3-0-2-dane-tlsa-records-with-le-certificates/7022

    plus various presentations listed at:

    https://github.com/baknu/DANE-for-SMTP/wiki/2.-Implementation-resources

--
	Viktor.