Re: [quicwg/base-drafts] Minor edits to amplification attack text (#1895)

[Martin Thomson <[e-mail]>]

martinthomson commented on this pull request.



>  verified through an address validation token (delivered via a Retry packet or
-a NEW_TOKEN frame) or by processing any message from the client encrypted using
-the Handshake keys.  This limit exists to mitigate amplification attacks.
+a NEW_TOKEN frame), by processing any message from the client encrypted using
+the Handshake keys, or by processing a TLS ClientHello with a valid
+pre-shared key.  This limit exists to mitigate amplification attacks.

That isn't right either.  Willingness to complete the handshake is a decision the server makes, and it does not need to be based on a test of the liveness of the remote address.

It's possible that TLS resumption (0-RTT or not) could bundle a token in the session ticket that allows for address validation, as we used to do.  But we no longer require that to happen.

A server can accept 0-RTT from an unvalidated client address.  The amount that you can send is limited, but it's not unreasonable.  The server doesn't need to send a certificate, so it can probably send some responses.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/quicwg/base-drafts/pull/1895#discussion_r227617645