Multipath QUIC

281 views
Skip to first unread message

徐孟尧

unread,
Feb 19, 2021, 10:59:11 PM2/19/21
to net-dev
Hi everyone, I am pretty interested in the multipath quic which chromium came up with. 
I checked out the draft and it's pretty illuminating.
I am just wondering that, is there any implementation for MPQUIC in the chromium project? I search the keywords but only find some code for log outputs. 
I did find some MPQUIC implementation in elder version of chromium, but all are deprecated a long time ago.
I am just thinking what is the current state of multipath quic implementation? Is it still being discussed?

David Schinazi

unread,
Feb 20, 2021, 1:40:27 PM2/20/21
to 徐孟尧, net-dev
Hi,

Multipath QUIC was removed from Chromium a few years ago because it did not yield positive results, and required significant complexity.

Thanks,
David

--
You received this message because you are subscribed to the Google Groups "net-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to net-dev+u...@chromium.org.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/net-dev/d3886c72-9748-4068-a17f-7197094f6e95n%40chromium.org.

Robin MARX

unread,
Feb 22, 2021, 8:28:18 AM2/22/21
to net-dev, David Schinazi, net-dev, qiutia...@gmail.com
Hello there,

While Multipath QUIC was removed from Google QUIC / Chromium's QUIC implementation, it is still actively discussed in the IETF and envisioned as a possible part of QUIC v2 (where v1 is the IETF QUIC setup being standardized now).

There are several proposals being tested in the IETF at the moment; for a nice overview and detailed musings, see Christian Huitema's recent blog posts on the topic [1].
All of this is highly experimental though, and multipath has not yet been adopted as an official item in the QUIC working group at the IETF (meaning it can take quite a while for it to actually become part of QUIC proper, if ever). 
The most advanced experimental IETF Multipath QUIC implementation at this point is probably pquic [2].

徐孟尧

unread,
Feb 22, 2021, 8:38:49 AM2/22/21
to net-dev, David Schinazi, net-dev, 徐孟尧
Thanks for the reply! 

徐孟尧

unread,
Feb 22, 2021, 8:39:10 AM2/22/21
to net-dev, Robin MARX, David Schinazi, net-dev, 徐孟尧
Thanks, will look into that!
Reply all
Reply to author
Forward
0 new messages