#DNS over #TLS authentication profiles draft have seen some modifications after exam by #IESG. Need some review (IMHO, it is done, and should be published). #IETF99
Alex Mayrhofer on the #DNS padding policy (to improve #privacy when encrypting with #TLS, which does not hide messages sizes). Draft near completion. #IETF99
The proposed padding policy comes from the excellent measurements done by Daniel K. Gilmore and presented at the #NDSS17 conference this year. We have facts to back the proposed policy ("pad to the next multiple of 128 bytes")
So you can have a #TLS server where an external adversary cannot tell if it provides #HTTPS or #DNS. (Does not work with #ALPN, though.) #privacy#IETF99
Shane Kerr c the "hallenges the "#DNS over #TCP is more costly than DNS over #UDP" story. UDP servers are hugely overprovisioned because of DoS attacks enabled by UDP. TCP may allow to decrease this overprovisioning. #IETF99