Re: [quicwg/base-drafts] One Retry/VN per UDP datagram (#2303)

[Kazuho Oku <[e-mail]>]

@marten-seemann Thank you for elaborating.

I think that we should be dispatching each QUIC datagram (instead of each QUIC packet) to the respective connections. That's how load balancers are expected to work, and what you are implementing is essentially a load balancer. The benefit of dispatching at the datagram level is that you'd have less things to do on the receiving thread.

> This wouldn't be a problem at all if a coalesced packet just had a few parts, as it's supposed to be. But since we're maximizing flexibility for no reason, an attacker might send me datagram consisting of around 70 packets, which I would all have to completely parse.

I understand your point on flexibility. However, in my view, the flexibility is not only how you coalesce, but the redundancy in all the long header fields. We have deliberately chosen the design because it is clear and easy to analyze (especially in regard to encryption). My preference goes to retaining the clarity by avoiding introducing restrictions.

-- 
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/2303#issuecomment-451816576