W3C home > Mailing lists > Public > ietf-http-wg@w3.org > April to June 2013

Re: HTTP/2 and TCP CWND

From: Yuchung Cheng <ycheng@google.com>
Date: Mon, 29 Apr 2013 09:41:36 -0700
Message-ID: <CAK6E8=dwNQxTK=j3xVa4X3nWdEJGGtU105hd0nVLSighrjTFZw@mail.gmail.com>
To: Mark Nottingham <mnot@mnot.net>
Cc: Martin Stiemerling <Martin.Stiemerling@neclab.eu>, "ietf-http-wg@w3.org Group" <ietf-http-wg@w3.org>, Lars Eggert <lars@netapp.com>, Gabriel Montenegro <Gabriel.Montenegro@microsoft.com>, "tcpm-chairs@ietf.org" <tcpm-chairs@ietf.org>
On Fri, Apr 26, 2013 at 11:12 PM, Mark Nottingham <mnot@mnot.net> wrote:
>
> On 26/04/2013, at 7:02 PM, Martin Stiemerling <Martin.Stiemerling@neclab.eu> wrote:
>
>> Hi all,
>>
>> I'm joining in late and have to second what Lars and other pointed out on that topic in respect of letting TCP do its job.
>>
>> On 04/15/2013 08:04 AM, Mark Nottingham wrote:
>>>> As for my opinion about what to do: I think we should delete this
>>>> TCP congestion window setting from HTTP/2.0.
>>>>
>>>> This is as out of scope as Iíve ever seen at the IETF. Modifying
>>>> TCP (by modifying its contract to upper layers such as HTTP, and by
>>>> modifying its state machine) is not something that can be done
>>>> outside of the Transport Area. Iím cc-ing Lars Eggert and Martin
>>>> Stiemerling (former and current Transport ADs), in case they have
>>>> additional comments or clarifications.
>>>
>>> I'd love to hear what Transport folks have to say on-list.
>>>
>>> I've also been talking with the Martin about having more Transport
>>> cross-fertilisation; we're hoping to spend some time on discussing
>>> relevant issues in Berlin.
>>
>> I wonder if we should have a HTTPbis presentation and discussion of the issue the HTTPbis WG faces to the TCPM group at the IETF in Berlin, sort of a joint session?

As long as the transport and http-bis started talking, that'd be a
great idea. So either is fine.

I also think some problems are beyond TCPM and should be discussed in
tsv-area. Today the congestion is often caused by connection
parallelism but that's not TCP's job to handle that. In HTTP/2 such a
problem still exist if the page contains resources from different
hosts.

>
>
> Makes sense to me. I'm happy to give some time in the HTTPbis session to this, or we could schedule a separate joint meeting.
>
> Cheers,
>
>
> --
> Mark Nottingham   http://www.mnot.net/
>
>
>
>
Received on Monday, 29 April 2013 16:42:24 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 11:11:12 UTC