Matt Corallo <[email protected]> writes:

> Am I incorrect in reading that both tls-option and tcpcrypt require a
> full three-way-handshake, and then the client is unable to send data for
> another round-trip as it must wait for another response from the server
> before session keys have been agreed upon?

In tcpcrypt it's half an extra half round trip.  So for protocols where
the server speaks first (including only protocols like SMTP, NNTP, IMAP,
etc., where the server sends a greeting), there is no penalty.  But for
client-speaks-first protocols (such as HTTP), there is an extra round
trip.

> Moving client pubkeys into SYN is slightly crazy, but would shave off a
> RTT before client's first send.

We'd all love to do that, but it is unlikely to happen before we get
large option support for SYNs.  So if you want to make this happen, your
best bet is to support something like tcpm-tcp-syn-ext-opt in the TCPM
working group and TCP-ENO in TCPINC.  The nice thing about TCP-ENO is
that it will naturally be able to take advantage of large SYN options as
soon as we have them.

David

_______________________________________________
Tcpinc mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/tcpinc

Reply via email to