Re: [quicwg/base-drafts] Long Header Packets and Routing Connection IDs (#2834)

[ianswett <[e-mail]>]

I think there are two obvious designs here, though there may be more:
 1) Change the CID routing based on packet type (Nick's suggestion above)
 2) Use 5-tuple routing for all long header packets(Because 5-tuple can't change during the handshake) and CID routing after that. 

If we care about the first case, I think we need to change the current design, though I'd be happy to be proven wrong.

One idea is to change to the new CID when sending packets of the next encryption level, since you have to parse the previous encryption level to start sending the next, but not change the current encryption level.

ie: if a client processes an Initial and Handshake, then all short header packets would be sent with the new Handshake CID, or if the CID was changed in Initial, then all Handshake and short header packets would be sent with that.

That leaves Retry.  I assume we want all Initial, Handshake, and 0-RTT packets to be sent with the new CID provided by the Retry, but I'd be curious what others think?

-- 
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/2834#issuecomment-504813305