Hi, Scott,

Thanks for this interesting information - it would be good to know as a
followup what the results are like with persistent TCP (following the RFC
recommendations), since the overhead of TCP connection setup for each
query/response before TLS resume is significant.

Related to assessment of encrypted DNS in the global Internet, I asked John
Todd how much usage of encrypted DNS the Quad9 recursive servers see, and
he wrote this (sent with permission):

Overall, across all locations, we see around 15% of our traffic is
encrypted in some form.  However, some locales (nations or regions) are
much higher.  Amsterdam, for instance, sees around 20% of all traffic
encrypted (note that AMS brings in traffic from much more distant locations
like Russia, China, and other privacy-challenged regions.)

And this rate is more a function of end users vs. forwarding caches.  We
have relatively few end users on our network - most of our traffic
(estimated >80%) comes from forwarding caches of some sort, many of them
very large.   So the fact that we have 15% encrypted is actually surprising
- that means a significant portion of our "end users" are turning on
encryption, since we see far fewer forwarding caches using encrypted
transport.

I don't have numbers to back up those statistics; we don't have a grant for
any of that work, so [further aspects of the data go] unexplored.


I see Quad9's findings as an indicator that DNS encryption has traction
with end users globally.

Regards,
Allison


On Thu, 12 Jun 2025 at 15:36, Hollenbeck, Scott <shollenbeck=
[email protected]> wrote:

> Earlier today I added text describing Verisign's RFC 9539 Experiment to
> GitHub:
>
>
> https://github.com/ietf-wg-dprive/9539-data/blob/main/Verisign's%20RFC%209539%20Experiment
>
> Scott
>
> _______________________________________________
> dns-privacy mailing list -- [email protected]
> To unsubscribe send an email to [email protected]
>
_______________________________________________
dns-privacy mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to