[{"author": "Mirja K\u00fchlewind", "text": "

if you are not presenting, please keep your video off

", "time": "2026-07-23T14:31:44.000Z"}, {"author": "Mirja K\u00fchlewind", "text": "

we will bring up the slides and hand over control to Johanna

", "time": "2026-07-23T14:32:02.000Z"}, {"author": "Mirja K\u00fchlewind", "text": "

you don't need to request slides.

", "time": "2026-07-23T14:32:11.000Z"}, {"author": "Mirja K\u00fchlewind", "text": "

are you co presenting? otherwise I will take your video off again

", "time": "2026-07-23T14:32:55.000Z"}, {"author": "Manu Gupta", "text": "

Thank you Mirja. Definitely.

", "time": "2026-07-23T14:34:16.000Z"}, {"author": "Mirja K\u00fchlewind", "text": "

thanks

", "time": "2026-07-23T14:34:27.000Z"}, {"author": "Mirja K\u00fchlewind", "text": "

If people can help note-taking that would be appreciated: https://notes.ietf.org/notes-ietf-126-atp

", "time": "2026-07-23T14:37:54.000Z"}, {"author": "Paul Carleton", "text": "

s/atp/maprg/
\nhttps://notes.ietf.org/notes-ietf-126-maprg

", "time": "2026-07-23T14:38:39.000Z"}, {"author": "Manu Gupta", "text": "

Thank you so much for attending the presentation. Wish to have your thoughts, feedback and pointers at manu@libp2p.io and johanna@libp2p.io

", "time": "2026-07-23T14:39:29.000Z"}, {"author": "Mirja K\u00fchlewind", "text": "

thanks, yes, would be great to put the maprg notes in the maprg notes!

", "time": "2026-07-23T14:45:01.000Z"}, {"author": "Manu Gupta", "text": "

Definitely, thank you Mirja. Appreciate it.

", "time": "2026-07-23T14:45:33.000Z"}, {"author": "Mirja K\u00fchlewind", "text": "

Manu, that was only a comment about me posting the wrong link. But also thanks for joining us today!

", "time": "2026-07-23T14:46:49.000Z"}, {"author": "Manu Gupta", "text": "

Yes, thank you Mirja. I did notice two links. Thanks for pointing to the right link: https://notes.ietf.org/notes-ietf-126-atp

\n

Also, thank you for the opportunity to present at MAPRG. Appreciate it.

", "time": "2026-07-23T14:53:07.000Z"}, {"author": "Mirja K\u00fchlewind", "text": "

The one with atp is the wrong link.

", "time": "2026-07-23T14:53:49.000Z"}, {"author": "Mirja K\u00fchlewind", "text": "

This is the right link: https://notes.ietf.org/notes-ietf-126-maprg

", "time": "2026-07-23T14:54:00.000Z"}, {"author": "Manu Gupta", "text": "

Great, thanks Mirja :)

", "time": "2026-07-23T14:55:05.000Z"}, {"author": "Jorge Silva", "text": "

Are the slides shared anywhere?

", "time": "2026-07-23T15:01:19.000Z"}, {"author": "Dave Plonka", "text": "

All the slides are linked on the IETF 126 agenda... look at the MAPRG timeslot, and there is a materials link on the right.

", "time": "2026-07-23T15:01:55.000Z"}, {"author": "Mirja K\u00fchlewind", "text": "

https://datatracker.ietf.org/meeting/126/session/maprg/

", "time": "2026-07-23T15:02:04.000Z"}, {"author": "Dave Plonka", "text": "

The pre-print for this work is available here (linked in the MAPRG agenda abstracts section): https://pure.mpg.de/rest/items/item_3705389_1/component/file_3705390/content

", "time": "2026-07-23T15:10:48.000Z"}, {"author": "Duane Wessels", "text": "

Johannes, have seen any evidence of TTL rewrites in protocols other than ICMP?

", "time": "2026-07-23T15:24:00.000Z"}, {"author": "Duane Wessels", "text": "

sorry, other than UDP?

", "time": "2026-07-23T15:25:11.000Z"}, {"author": "Mirja K\u00fchlewind", "text": "

@meetecho any chance you can zoom in for us for this talk?

", "time": "2026-07-23T15:25:16.000Z"}, {"author": "Lorenzo Miniero", "text": "

@Mirja K\u00fchlewind like this? Or did you mean the audience?

", "time": "2026-07-23T15:25:57.000Z"}, {"author": "Mirja K\u00fchlewind", "text": "

the slides

", "time": "2026-07-23T15:26:08.000Z"}, {"author": "Johannes Zirngibl", "text": "

Yes, we also tested icmp echo messages to check that it is independent from very small TTL values in the probes and saw it for TCP based traceroutes as well

", "time": "2026-07-23T15:26:33.000Z"}, {"author": "Dave Plonka", "text": "

If you want to get a closer look at Nalini's slides b/c of the vertical format, they are here: https://datatracker.ietf.org/meeting/126/materials/slides-126-maprg-measuring-deployment-characteristics-of-post-quantum-tls-authentication-mechanisms-00

", "time": "2026-07-23T15:26:37.000Z"}, {"author": "Mirja K\u00fchlewind", "text": "

sorry for the slides format

", "time": "2026-07-23T15:26:38.000Z"}, {"author": "Lorenzo Miniero", "text": "

Ah then I'm afraid not :sad:

", "time": "2026-07-23T15:26:54.000Z"}, {"author": "Mirja K\u00fchlewind", "text": "

@meetecho okay, just checking. thx!

", "time": "2026-07-23T15:27:06.000Z"}, {"author": "Dave Plonka", "text": "

Robert's paper is here, btw: https://robert-richter.me/pdfs/starlink-dns-dynamics.pdf

\n

Also linked in the MAPRG agenda, in the abstracts section.

", "time": "2026-07-23T15:36:14.000Z"}, {"author": "Valentin Go\u0219u", "text": "

@Robert Richter For DoE (and even TCP) measurements, it would be interesting if you also measured latency over a persistent connection (without the initial handshake delay)

", "time": "2026-07-23T15:42:53.000Z"}, {"author": "Dave Plonka", "text": "

The paper Newton is presenting can be found here, unencumbered by paywall: https://vaibhavbajpai.com/documents/papers/proceedings/2026-pam-doq-doh3.pdf

", "time": "2026-07-23T15:45:46.000Z"}, {"author": "Stephen Farrell", "text": "

remind me: is DoQ with 0-RTT vulnerable to replay if sent to an anycast instance?

", "time": "2026-07-23T15:47:10.000Z"}, {"author": "Benjamin Schwartz", "text": "

@stephen Complicated. The answer is something like SHOULD NOT. But the response is encrypted so you don't learn much from the replay.

", "time": "2026-07-23T15:48:23.000Z"}, {"author": "Jason Livingood", "text": "

those RTTs still seem high

", "time": "2026-07-23T15:48:32.000Z"}, {"author": "Stephen Farrell", "text": "

unless you can also see the other side of the anycast instance

", "time": "2026-07-23T15:49:03.000Z"}, {"author": "Jason Livingood", "text": "

Oof - these PLTs show how bloated the avg site has gotten

", "time": "2026-07-23T15:50:32.000Z"}, {"author": "Erik Nygren", "text": "

@stephen: right, if you can see both sides you might be able to extract some things if the resolver is using Do53 to the authorities, but I suspect the correlation you can already do in that position is dominant even without 0RTT replay. Since these are mostly idempotent (other than for the case you raise) it's unclear how much of a problem it is in practice.

", "time": "2026-07-23T15:50:44.000Z"}, {"author": "Lucas Pardue", "text": "

https://httparchive.org/reports/page-weight

", "time": "2026-07-23T15:51:15.000Z"}, {"author": "Jason Livingood", "text": "

aka \"the web is so bloated, users won't notice slow DNS QRTs\" ;-)

", "time": "2026-07-23T15:51:38.000Z"}, {"author": "Stephen Farrell", "text": "

IIRC anycast DNS was the canonical example of when to not do 0-RTT

", "time": "2026-07-23T15:52:09.000Z"}, {"author": "Marco Munizaga", "text": "

Very different numbers of page weight from httparchive and the presenter's slides (1.2 mb vs 3mb)?

", "time": "2026-07-23T15:52:44.000Z"}, {"author": "Dave Plonka", "text": "

We're going to close the queue after Erik Nygren.

", "time": "2026-07-23T15:55:32.000Z"}, {"author": "Benjamin Schwartz", "text": "

@Stephen In TLS 1.3, servers SHOULD prevent 0-RTT replay. For large anycast networks, this is presumably done by partitioning resumption for each local server pool, and having a replay prevention database in each location.

", "time": "2026-07-23T15:55:38.000Z"}, {"author": "Lars Eggert", "text": "

Forgot to mention on the mic that for us, DoH resolves FASTER than Do53

", "time": "2026-07-23T15:55:57.000Z"}, {"author": "Lars Eggert", "text": "

At - I think - P75

", "time": "2026-07-23T15:56:06.000Z"}, {"author": "Stephen Farrell", "text": "

@ben: have any of 'em siad that? if not, I'd be concerned if stubs were using 0-RTT (not emergency concerned but concerned)

", "time": "2026-07-23T15:56:37.000Z"}, {"author": "Stephen Farrell", "text": "

mind you, I've been concerned about 0-RR from the start but the sky hasn't fallen yet:-)

", "time": "2026-07-23T15:58:13.000Z"}, {"author": "Benjamin Schwartz", "text": "

I believe this is a common practice in general for TLS. Globally resumable session tickets require globally distributed secret keys, which is a very weak security posture.

", "time": "2026-07-23T15:58:26.000Z"}, {"author": "Jason Livingood", "text": "

For the presenter (since queue is closed) - when as an operator we optimize for RTT and delay in general, always look at p95 or p99 not mean or median as those will mislead and sometimes give the impression all is well. So p95 or p99.

", "time": "2026-07-23T15:58:58.000Z"}, {"author": "Lars Eggert", "text": "

https://performance.mozilla.org/networking.html has TRR DoH performance numbers

", "time": "2026-07-23T15:59:05.000Z"}, {"author": "Stephen Farrell", "text": "

@lars: does FF use 0-RTT for DoH?

", "time": "2026-07-23T15:59:23.000Z"}, {"author": "Tim Chown", "text": "

I was just going to suggest he speaks to the web tester team about happy-eyeballs.net. they're in Germany too I recall.

", "time": "2026-07-23T15:59:38.000Z"}, {"author": "Dave Plonka", "text": "

Shyam's paper is available here, from TMA 2026: https://ris.utwente.nl/ws/portalfiles/portal/538780291/Exposed_BGP_routers.pdf

", "time": "2026-07-23T15:59:55.000Z"}, {"author": "Dave Plonka", "text": "

good idea, @tim.

", "time": "2026-07-23T16:00:35.000Z"}, {"author": "Lars Eggert", "text": "

@stephen: Yes, but also the connection is very long lived

", "time": "2026-07-23T16:00:50.000Z"}, {"author": "Stephen Farrell", "text": "

ack

", "time": "2026-07-23T16:01:00.000Z"}, {"author": "Lucas Pardue", "text": "

should DNS over HTTPS be using the QUERY method rther than POST?

", "time": "2026-07-23T16:01:30.000Z"}, {"author": "Lars Eggert", "text": "

We recently updated https://wiki.mozilla.org/Security/DOH-resolver-policy and one change was to require idle timeouts > 180s

", "time": "2026-07-23T16:01:58.000Z"}, {"author": "Lars Eggert", "text": "

@lucas Moz switched to GET recently

", "time": "2026-07-23T16:02:33.000Z"}, {"author": "Dave Plonka", "text": "

A paper with the work Ebrima is reporting on is here: https://dl.ifip.org/db/conf/cnsm/cnsm2025/1571167978.pdf

", "time": "2026-07-23T16:14:08.000Z"}, {"author": "Dave Plonka", "text": "

Also, a shorter blog version here:
\nhttps://blog.apnic.net/2026/04/16/noisy-routers-investigating-the-make-up-of-route-collector-data/

", "time": "2026-07-23T16:17:50.000Z"}, {"author": "Dave Plonka", "text": "

For those that might be interested, the voluminous RouteViews data is mirrored to here: https://osg-htc.org/services/osdf/data?repository=routeviews

\n

OSDF and partners are looking for tutorials/examples of projects using it via OSDF, or are interested in streaming access. Please let me know if you are!

", "time": "2026-07-23T16:22:30.000Z"}, {"author": "Manu Gupta", "text": "

Thanks for sharing, Dave. This looks quite relevant for the libp2p/IPFS ecosystem. The RouteViews dataset could be valuable for correlating Internet routing dynamics with P2P connectivity, DHT behavior, and network resilience. The streaming access is especially interesting for observability and measurement research. We'd be happy to learn more and explore potential collaboration.

", "time": "2026-07-23T16:28:38.000Z"}, {"author": "Mirja K\u00fchlewind", "text": "

Please feel free to use the mailing list for further discussion/questions!

", "time": "2026-07-23T16:29:12.000Z"}, {"author": "Mirja K\u00fchlewind", "text": "

(maprg@ietf.org)

", "time": "2026-07-23T16:29:29.000Z"}, {"author": "Manu Gupta", "text": "

Great, thanks Mirja :)

", "time": "2026-07-23T16:29:32.000Z"}, {"author": "Tim Chown", "text": "

Ipv6matrix.com has the data just presented going back maybe 15 years

", "time": "2026-07-23T16:29:37.000Z"}, {"author": "Marco Munizaga", "text": "

I'm suprised .jp domains are so low in IPv6 adoption?

", "time": "2026-07-23T16:30:28.000Z"}, {"author": "Manu Gupta", "text": "

Indeed, Marco :)

", "time": "2026-07-23T16:30:52.000Z"}]