Re: http/2 prioritization/fairness bug with proxies

Nico Williams <nico@cryptonector.com> Tue, 12 February 2013 00:01 UTC

Return-Path: <ietf-http-wg-request@listhub.w3.org>
X-Original-To: ietfarch-httpbisa-archive-bis2Juki@ietfa.amsl.com
Delivered-To: ietfarch-httpbisa-archive-bis2Juki@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 8767B21F8976 for <ietfarch-httpbisa-archive-bis2Juki@ietfa.amsl.com>; Mon, 11 Feb 2013 16:01:32 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -7.888
X-Spam-Level:
X-Spam-Status: No, score=-7.888 tagged_above=-999 required=5 tests=[AWL=1.789, BAYES_00=-2.599, FM_FORGED_GMAIL=0.622, MIME_8BIT_HEADER=0.3, RCVD_IN_DNSWL_HI=-8]
Received: from mail.ietf.org ([64.170.98.30]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id GBcsGhARXjPF for <ietfarch-httpbisa-archive-bis2Juki@ietfa.amsl.com>; Mon, 11 Feb 2013 16:01:31 -0800 (PST)
Received: from frink.w3.org (frink.w3.org [128.30.52.56]) by ietfa.amsl.com (Postfix) with ESMTP id D84B521F8848 for <httpbisa-archive-bis2Juki@lists.ietf.org>; Mon, 11 Feb 2013 16:01:31 -0800 (PST)
Received: from lists by frink.w3.org with local (Exim 4.72) (envelope-from <ietf-http-wg-request@listhub.w3.org>) id 1U53II-0008J5-6i for ietf-http-wg-dist@listhub.w3.org; Tue, 12 Feb 2013 00:00:18 +0000
Resent-Date: Tue, 12 Feb 2013 00:00:18 +0000
Resent-Message-Id: <E1U53II-0008J5-6i@frink.w3.org>
Received: from lisa.w3.org ([128.30.52.41]) by frink.w3.org with esmtp (Exim 4.72) (envelope-from <nico@cryptonector.com>) id 1U53IA-00083j-Ai for ietf-http-wg@listhub.w3.org; Tue, 12 Feb 2013 00:00:10 +0000
Received: from caiajhbdccac.dreamhost.com ([208.97.132.202] helo=homiemail-a85.g.dreamhost.com) by lisa.w3.org with esmtp (Exim 4.72) (envelope-from <nico@cryptonector.com>) id 1U53I9-0001Mp-9W for ietf-http-wg@w3.org; Tue, 12 Feb 2013 00:00:10 +0000
Received: from homiemail-a85.g.dreamhost.com (localhost [127.0.0.1]) by homiemail-a85.g.dreamhost.com (Postfix) with ESMTP id 7B2C8BC04D for <ietf-http-wg@w3.org>; Mon, 11 Feb 2013 15:59:46 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha1; c=relaxed; d=cryptonector.com; h= mime-version:in-reply-to:references:date:message-id:subject:from :to:cc:content-type:content-transfer-encoding; s= cryptonector.com; bh=w+98BzNBxIfJ837YnNa6ShcnpeQ=; b=PS+JTPkCK1t /OCopoDRp+TMMlQ3U5Oo/HEjt9KniXpoom/PbxNxRg0I1hxRu/VO33Sh0iwndVkT 0/2s3iZOTSWJqi0rRWyYfDU8eUE6G6w7x/t3tAqDNIEnPMVSP05hiI8WCZC+NE4a C612objm/g5UjW+9fn4SaJC0PYFelqCE=
Received: from mail-wg0-f51.google.com (mail-wg0-f51.google.com [74.125.82.51]) (using TLSv1 with cipher RC4-SHA (128/128 bits)) (No client certificate requested) (Authenticated sender: nico@cryptonector.com) by homiemail-a85.g.dreamhost.com (Postfix) with ESMTPSA id 29400BC042 for <ietf-http-wg@w3.org>; Mon, 11 Feb 2013 15:59:46 -0800 (PST)
Received: by mail-wg0-f51.google.com with SMTP id 8so5083034wgl.18 for <ietf-http-wg@w3.org>; Mon, 11 Feb 2013 15:59:44 -0800 (PST)
MIME-Version: 1.0
X-Received: by 10.180.8.130 with SMTP id r2mr6080758wia.28.1360627184892; Mon, 11 Feb 2013 15:59:44 -0800 (PST)
Received: by 10.217.39.133 with HTTP; Mon, 11 Feb 2013 15:59:44 -0800 (PST)
In-Reply-To: <CAA4WUYg+evqEMdiYSv+EZ7668eCq_dwqKiYmA4Lq-xZkoD_9Fw@mail.gmail.com>
References: <CAA4WUYjiBZpShKKFfHQnixc94aOLrck0oR4ykARB=hF5h8nkfA@mail.gmail.com> <3430.1359961022@critter.freebsd.dk> <510F72CE.8030003@treenet.co.nz> <CAA4WUYiBJrLjM0-vurFOuJfUaabXtK=W8N5z28yshSfrvD9crg@mail.gmail.com> <1516.1360002578@critter.freebsd.dk> <42A54D15-0AA3-4172-94F7-E94C86E84D7F@niven-jenkins.co.uk> <2346.1360010079@critter.freebsd.dk> <CAA4WUYiyu+JvFuKooqa4xVdCJP=Mngu9dgHjhH99_SEac1kCZQ@mail.gmail.com> <CABaLYCtX04se2BJ0c1yCYkwH2xvkcETdu7Pe+B8fy=DJrouo6Q@mail.gmail.com> <CAA4WUYitNR0js+5m5RHfLp-7=k-mfTUKDcayW-uzJwTgOgMyYQ@mail.gmail.com> <CABkgnnUYKDe4rZ0ZpqASkwDF2Foa_ni1rEFJH1P03k1Bv_NCog@mail.gmail.com> <CAA4WUYg+evqEMdiYSv+EZ7668eCq_dwqKiYmA4Lq-xZkoD_9Fw@mail.gmail.com>
Date: Mon, 11 Feb 2013 17:59:44 -0600
Message-ID: <CAK3OfOi7So1KWXdfv+UEqKAr-o1TaXiZmSmJx61WFR3J0zW_JA@mail.gmail.com>
From: Nico Williams <nico@cryptonector.com>
To: "William Chan (陈智昌)" <willchan@chromium.org>
Cc: Martin Thomson <martin.thomson@gmail.com>, Mike Belshe <mike@belshe.com>, Poul-Henning Kamp <phk@phk.freebsd.dk>, Ben Niven-Jenkins <ben@niven-jenkins.co.uk>, Amos Jeffries <squid3@treenet.co.nz>, HTTP Working Group <ietf-http-wg@w3.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Received-SPF: none client-ip=208.97.132.202; envelope-from=nico@cryptonector.com; helo=homiemail-a85.g.dreamhost.com
X-W3C-Hub-Spam-Status: No, score=-3.4
X-W3C-Hub-Spam-Report: AWL=-3.316, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, RCVD_IN_DNSWL_NONE=-0.0001
X-W3C-Scan-Sig: lisa.w3.org 1U53I9-0001Mp-9W 70207ff749cf4e7f07b0bbbb9a5e7a6e
X-Original-To: ietf-http-wg@w3.org
Subject: Re: http/2 prioritization/fairness bug with proxies
Archived-At: <http://www.w3.org/mid/CAK3OfOi7So1KWXdfv+UEqKAr-o1TaXiZmSmJx61WFR3J0zW_JA@mail.gmail.com>
Resent-From: ietf-http-wg@w3.org
X-Mailing-List: <ietf-http-wg@w3.org> archive/latest/16579
X-Loop: ietf-http-wg@w3.org
Resent-Sender: ietf-http-wg-request@w3.org
Precedence: list
List-Id: <ietf-http-wg.w3.org>
List-Help: <http://www.w3.org/Mail/>
List-Post: <mailto:ietf-http-wg@w3.org>
List-Unsubscribe: <mailto:ietf-http-wg-request@w3.org?subject=unsubscribe>

On Mon, Feb 11, 2013 at 4:46 PM, William Chan (陈智昌)
<willchan@chromium.org> wrote:
> Theoretically possible is one thing. But the moment we get into the game of
> trying to carve up portions of BDP via per-stream flow control windows for
> prioritization purposes in normal operation (as opposed to just trying to
> make reasonable progress under excessive load), I think we're in trouble,
> and likely to get into performance issues due to poor implementations. As
> I've stated before, I hope most implementations (and believe we should add
> recommendations for this behavior) only use flow control (if they use it at
> all, which hopefully they don't because it's hard) for maintaining
> reasonable buffer sizes.

Right.  Don't duplicate the SSHv2 handbrake (Peter Gutmann's term) in HTTP/2.0.

Use percentages of BDP on the sender side.  Have the receiver send
control frames indicating the rate at which it's receiving to help
estimate BDP, or ask TCP.  But do not flow control.

Another possibility is to have the sender (or a proxy) use
per-priority TCP connections.

Nico
--