Re: [quicwg/base-drafts] rate-limiting of CID issuance needs to be allowed (#2436)

[Kazuho Oku <[e-mail]>]

> I am concerned about non-HTTP uses cases, and for that matter also about mixing the layers. Transport should not rely on GOAWAY, and HTTP layer should be not concerned with refreshing CID's.

Fair enough. Though my point is that _for HTTP/3_ (which happens to be the primary target of QUIC v1) just capping the number of CIDs is totally fine. The proposal does not mandate such behavior. It's merely clarifying that such behavior is permissible.

When you are to define an application protocol for QUIC v1 that requires the connections to survive lots of migrations, you would be required to clarify that the servers should support that, regardless of the outcome of this issue. That is because, as stated in [my previous comment](
https://github.com/quicwg/base-drafts/issues/2436#issuecomment-461919062), migration is an optional feature of QUIC v1. Running out of CID is no worse than connecting to a server that does not support migration in terms of the transport spec.

To put it another way, while I can understand your concern, I think it's orthogonal to this issue.

-- 
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/2436#issuecomment-461997482