Re: [quicwg/base-drafts] Request to Retire Locally Issued CIDs (#2769)

[Marten Seemann <[e-mail]>]

I agree that atomicity is important here, if we want to avoid the situation where all connections IDs are retired at the same time. Adding the field to the NEW_CONNECTION_ID frame as @kazuho suggested seems to be the most straightforward way to do so.

I'm not a big fan of the 3 PTO requirement here. I don't see why it shouldn't be possible to retire the connections immediately when receiving the frame. There's a DoS risk here: if an attacker delays a packet with a connection ID that was retired afterwards, it can cause either a stateless reset or even a PROTOCOL_VIOLATION. The better way to enforce the retiring of the connection IDs is to look at packet numbers: The old connection IDs MUST NOT be used in packets with higher packet numbers than the packet that acknowledged the NEW_CONNECTION_ID frame.

-- 
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/2769#issuecomment-499375012