Re: [quicwg/base-drafts] The method of identifying "the same server" (#3155)

[Nick Harper <[e-mail]>]

I agree that a or b is not restrictive enough based on the example given above.

I wonder if c/d/e are also not restrictive enough though. Consider example.com hosted on multiple CDNs (where QUIC is terminated by the CDN). A client makes an initial connection to example.com (which gets served by CDN A) and receives a NEW_TOKEN token. Later, this client looks up example.com, gets directed to CDN B, and sends the token it received on the connection to CDN A. I wouldn't expect CDN B to be able to process that token.

We could also decide that it's not worth solving the case above. Being overly restrictive in scoping tokens could mean we don't save tokens often enough (resulting in connections being established with no token when we could have used a token), which might be worse than occasionally sending the wrong token to a server.

-- 
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/issues/3155#issuecomment-547098249