
From nobody Mon Jun  1 02:47:15 2020
Return-Path: <fabio.bulgarella@guest.telecomitalia.it>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 4287F3A0EBB for <quic@ietfa.amsl.com>; Mon,  1 Jun 2020 02:47:11 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.501
X-Spam-Level: 
X-Spam-Status: No, score=-1.501 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, GB_ABOUTYOU=0.5, RCVD_IN_MSPIKE_H2=-0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=no autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=guest.telecomitalia.it
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 2RcRkm6IvbNF for <quic@ietfa.amsl.com>; Mon,  1 Jun 2020 02:47:07 -0700 (PDT)
Received: from mx07.telecomitalia.it (mx07.telecomitalia.it [156.54.232.23]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 301D33A0EBC for <quic@ietf.org>; Mon,  1 Jun 2020 02:47:06 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; d=guest.telecomitalia.it; s=selector1; c=relaxed/relaxed;  q=dns/txt; i=@guest.telecomitalia.it; t=1591004823; x=2454918423; h=MIME-Version:From:Date:Message-ID:Subject:To; bh=WdrKefF1sfz8oL6fyB1MusvTFNs9WLVwDa23ojcNjG4=; b=My7ONUuTVJowjuoZnPonisHhK7AMHK4e7rzFxk27sz4ng4CSpXl6VvQxeLxDLxMZ JXiAl0NXrOQSpfuEfx5zt0x3cOUxAksAwM8dmqm7jqyEOHRZelcRBpFr/I81Bn0h mjHmmCwWJydLMaC+9ydTmA3wfQRmqZWFXnnKMqOHBgE27w2GxvPSjBh9bEuZRmy/ pWE0bWtWIk6hRrvz8Z5Hp/Zx9ekdVNG5IyUre13WObpMlrwNOn78FvNwuc+x0H/o CIfOMlOKYhKUaeIj7K6GC5MizKz4fih/gaUIGs9V0YmHxeISTjH78Adi7NRMUC5u klkxJlQgksJkfkWPUIPAYw==;
X-AuditID: 0a5a2d17-4a7ff700000048f6-cf-5ed4ce969b6e
Received: from TELMBXC12BA020.telecomitalia.local ( [10.90.43.46]) by mx07.telecomitalia.it () with SMTP id 54.87.18678.69EC4DE5; Mon,  1 Jun 2020 11:47:03 +0200 (CEST)
From: "Bulgarella Fabio (Guest)" <fabio.bulgarella@guest.telecomitalia.it>
To: Dmitri Tikhonov <dtikhonov@litespeedtech.com>
CC: "Lubashev, Igor" <ilubashe@akamai.com>, Ian Swett <ianswett@google.com>, Cociglio Mauro <mauro.cociglio@telecomitalia.it>, "isabelle.hamchaoui@orange.com" <isabelle.hamchaoui@orange.com>, "alexandre.ferrieux@orange.com" <alexandre.ferrieux@orange.com>, "quic@ietf.org" <quic@ietf.org>, "IETF IPPM WG (ippm@ietf.org)" <ippm@ietf.org>, Marcus Ihlar <marcus.ihlar@ericsson.com>
Subject: Re: [EXT] Re: [ippm] New QUIC Packet Loss Measurement (draft-cfb-ippm-spinbit-measurements)
Thread-Topic: [EXT] Re: [ippm] New QUIC Packet Loss Measurement (draft-cfb-ippm-spinbit-measurements)
Thread-Index: AQHWF/zfNwOq9+2pFU28gxDKDgXEdaiFgojx///ut4CAAT/o4YAGTsMAgDbA/14=
Date: Mon, 1 Jun 2020 09:47:01 +0000
Message-ID: <1591004822041.41433@guest.telecomitalia.it>
References: <3ca3b5aae01d4650a3451639268b3f1e@TELMBXD14BA020.telecomitalia.local> <CAKcm_gMEELBizN_h5+s3Ow0LKXEgTRGg+-AqzJMZXVBDwQcDLA@mail.gmail.com> <6b9e74ac94114d28ae4a66f1e9625ebd@usma1ex-dag1mb6.msg.corp.akamai.com> <1587582698016.72632@guest.telecomitalia.it> <20200422200530.GA11856@lubuntu> <1587648443587.63341@guest.telecomitalia.it>, <20200427152957.GC24088@lubuntu>
In-Reply-To: <20200427152957.GC24088@lubuntu>
Accept-Language: it-IT, en-US
Content-Language: it-IT
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
x-ms-exchange-transport-fromentityheader: Hosted
x-originating-ip: [188.219.222.222]
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFlrOIsWRmVeSWpSXmKPExsXCFaWtpzv93JU4g+fHmCyW/ZC3+DKBzeLn vZ2sFk0NK5gteh68Y7a4dPAju8XBl69YLXoWcDtweEw+soDZ49fXq2weCzaVeixZ8pPJ48m2 48weLc9OsgWwRXHZpKTmZJalFunbJXBlTHscXLDNuGL7o4lMDYwbNbsYOTkkBEwkvh5dzdLF yMUhJLCKUWLK7ofsXYwcHGwCXhKz7yiDmCIChhL79iaDlDALtDBLzD19kxmkV1ggTeLB1/mM ILaIQLrE0VN9rBD1fhI/FvGAhFkEVCROT3gFVs4rYCHxdhpIOciqRmaJGU2v2EASnAK6Ertu XWUCsRkFZCUm7F4ENpNZQFzixfQT7BB3Ckgs2XOeGcIWlXj5+B8rhG0gsXXpPhYIW1li9Z1r rBC9ehI3pk5hg7C1JZYtfA11hKDEyZlPwOqFgH7/sPIcywRGsVlI1s1C0j4LSfssJO0LGFlW MYrmVhiY65Wk5qQm5+dmliTmZCbqZZZsYgTHpa74DsYJG97oHWJk4mA8xCjBwawkwjtZ/VKc EG9KYmVValF+fFFpTmrxIUZpDhYlcV63A1fihATSE0tSs1NTC1KLYLJMHJxSDUySLo+iP161 Sd43TYRZ/AaHaXB9rsjdtLLENlehaYf8TGfWld06zaKVt9p2za8l0z6562wT0b2fwdK1QPC+ k+jGZIYnNy8c5Chn3dtVo1S35cV6+dnzJB9PkKxR8WzKFFFm/3Bm8o+LG8I+Tty8TzbxceL8 dTyFN48LPLTy1Qrr7G+Z7L7RRdv03v9sXs/025NlIl8f5vm8e/O+C1X71zcKVHrP+Odn5l50 m49l6Vz5REbNR71ZKrO3lm/6PrGBweLauVt7mX78Mq4PK2dZyLdztza31z7bEuFPS5eIbVkX OP+gtT73iePTy415XW5nFf58qOulXru0K9v17Nr9P0KmSR3aVrrnYl9+3ZZvTJpKLMUZiYZa zEXFiQDf1bqwOgMAAA==
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/4WVPycDvA6uIxdrPNorhsPT5MzM>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 01 Jun 2020 09:47:11 -0000

Hello Dmitri,=0A=
sorry for the late reply. We have just replied to Igor concerns in a previo=
us mail proposing a comparison session.=0A=
Our proposal is to implement both solutions on a common implementation of t=
he QUIC protocol so that the same data flow can be marked using both techni=
ques (i.e., Qbit, Rbit and Lbit - borrowing the third most significant bit =
occupied by the spinbit) and analyzed using the two observers.=0A=
Let us know what you think, thanks.=0A=
=0A=
Here, just few thoughts about your last considerations.=0A=
=0A=
=0A=
> Consider the following setup:=0A=
>=0A=
>    Client ... (OpA) ... (OpB) ... (OpC) ... Server=0A=
>=0A=
> We are OpB, and Client's and Server's lawyers are banging at our door=0A=
> with a legally binding SLA signed with our blood.  What do you think is=
=0A=
> a realistic way of getting rid of the situation:=0A=
>=0A=
>    a) Here are captures that prove that loss is in OpC, or in your=0A=
>       premises, Mr Server.=0A=
>=0A=
>    b) Here are captures that have no losses in OpA or OpB.  Have a=0A=
>       good day.=0A=
>=0A=
> Since a unidirectional QR capture does not include any evidence of=0A=
> loss downstream, it would be dismissed by the Server lawyers as just=0A=
> not capturing the right traffic.  An equivalent packet dump with TCP=0A=
> would include evidence of loss in the form of packet retransmission,=0A=
> and so would QL scheme.=0A=
>=0A=
> Simultaneous captures by different organizations are often hard to=0A=
> coordinate, especially for problems that do not reproduce all the time.=
=0A=
> At any given point of the investigation, you depend only on your own=0A=
> local resources and need to be able to provide evidence that makes sense=
=0A=
> to everybody.=0A=
=0A=
=0A=
Regarding this point, with respect to a signed SLA, we believe that there a=
re no big differences between asserting that the problem is not in our doma=
in or asserting that the problem is outside our domain. In any case, your s=
ystem is still unable to locate the leak outside your domain: you cannot sa=
y if it is in the OpC or in any other operator in the queue (unless you cal=
l OpC and ask for place a probe in their domain).  Furthermore, referring t=
o your example, considering that the OpC also does not experience any probl=
ems, with the QR method we can measure E2E on the opposite side without hav=
ing a probe physically placed on that side. So you have more information to=
 use in the event of a legal dispute.=0A=
=0A=
=0A=
> How common is "not uncommon:" 90%? 50%?  Here is a study that shows how=
=0A=
> common symmetric flows are: out of six observations of four different=0A=
> Internet links, the percentage of symmetric flows was 57%, 7%, 2%, 3%,=0A=
> 2%, and 3%:=0A=
>=0A=
>  https://www.caida.org/research/traffic-analysis/asymmetry/=0A=
>=0A=
> The way routing works is asymmetric in essence, as the destination decide=
s=0A=
> what it announces, which leads to the way to reach it.  When you revert=
=0A=
> the arrow, the source cannot but just follow the routes.  The result is=
=0A=
> a high prevalence of asymmetric routing.=0A=
=0A=
=0A=
If observation is possible on both sides, both the QL and QR methods work b=
est. If, on the other hand, only one direction can be observed, each of the=
 two solutions has its pros and cons: you see everything on the observed si=
de but you are totally blind on the opposite side. We only see our domain o=
n the side under observation and also E2E on the opposite side. What is bet=
ter? Any contribution in this sense from the working groups is welcome.=0A=
=0A=
=0A=
Best regards,=0A=
Fabio B.=0A=
=0A=
=0A=
=0A=
________________________________________=0A=
Da: Dmitri Tikhonov <dtikhonov@litespeedtech.com>=0A=
Inviato: luned=EC 27 aprile 2020 17:29=0A=
A: Bulgarella Fabio (Guest)=0A=
Cc: Lubashev, Igor; Ian Swett; Cociglio Mauro; isabelle.hamchaoui@orange.co=
m; Riccardo Sisto; alexandre.ferrieux@orange.com; quic@ietf.org; IETF IPPM =
WG (ippm@ietf.org); Nilo Massimo=0A=
Oggetto: Re: [EXT] Re: [ippm] New QUIC Packet Loss Measurement (draft-cfb-i=
ppm-spinbit-measurements)=0A=
=0A=
Hello Fabio,=0A=
=0A=
On Thu, Apr 23, 2020 at 01:27:23PM +0000, Bulgarella Fabio (Guest) wrote:=
=0A=
> > Do you not agree that locating loss source is one of the probe's goals?=
=0A=
>=0A=
> It is more interesting to know if a problem occurs in our domain or=0A=
> outside. And this is done by the Qbit (With 2 probes).=0A=
=0A=
Consider the following setup:=0A=
=0A=
    Client ... (OpA) ... (OpB) ... (OpC) ... Server=0A=
=0A=
We are OpB, and Client's and Server's lawyers are banging at our door=0A=
with a legally binding SLA signed with our blood.  What do you think is=0A=
a realistic way of getting rid of the situation:=0A=
=0A=
    a) Here are captures that prove that loss is in OpC, or in your=0A=
       premises, Mr Server.=0A=
=0A=
    b) Here are captures that have no losses in OpA or OpB.  Have a=0A=
       good day.=0A=
=0A=
Since a unidirectional QR capture does not include any evidence of=0A=
loss downstream, it would be dismissed by the Server lawyers as just=0A=
not capturing the right traffic.  An equivalent packet dump with TCP=0A=
would include evidence of loss in the form of packet retransmission,=0A=
and so would QL scheme.=0A=
=0A=
Simultaneous captures by different organizations are often hard to=0A=
coordinate, especially for problems that do not reproduce all the time.=0A=
At any given point of the investigation, you depend only on your own=0A=
local resources and need to be able to provide evidence that makes sense=0A=
to everybody.=0A=
=0A=
> However, it is not uncommon to be able to intercept the whole traffic=0A=
> of a connection by observing a single point.  In this case, all the=0A=
> necessary measurements can be made.=0A=
=0A=
How common is "not uncommon:" 90%? 50%?  Here is a study that shows how=0A=
common symmetric flows are: out of six observations of four different=0A=
Internet links, the percentage of symmetric flows was 57%, 7%, 2%, 3%,=0A=
2%, and 3%:=0A=
=0A=
  https://www.caida.org/research/traffic-analysis/asymmetry/=0A=
=0A=
The way routing works is asymmetric in essence, as the destination decides=
=0A=
what it announces, which leads to the way to reach it.  When you revert=0A=
the arrow, the source cannot but just follow the routes.  The result is=0A=
a high prevalence of asymmetric routing.=0A=
=0A=
  - Dmitri.=0A=
=0A=


From nobody Mon Jun  1 03:47:29 2020
Return-Path: <lars@eggert.org>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id F248C3A0F90 for <quic@ietfa.amsl.com>; Mon,  1 Jun 2020 03:47:20 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.099
X-Spam-Level: 
X-Spam-Status: No, score=-2.099 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=eggert.org
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id hXtjzcW9eWlC for <quic@ietfa.amsl.com>; Mon,  1 Jun 2020 03:47:19 -0700 (PDT)
Received: from mail.eggert.org (mail.eggert.org [91.190.195.94]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id A51043A0F32 for <quic@ietf.org>; Mon,  1 Jun 2020 03:47:16 -0700 (PDT)
Received: from [IPv6:2a00:ac00:0:35:1877:3bc3:26eb:bf3a] (unknown [IPv6:2a00:ac00:0:35:1877:3bc3:26eb:bf3a]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.eggert.org (Postfix) with ESMTPSA id EEA4C616E99 for <quic@ietf.org>; Mon,  1 Jun 2020 13:47:07 +0300 (EEST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=eggert.org; s=dkim; t=1591008428; bh=Lf2AgvAnC5xE/MlO99QsHNt+kMVGJH+gj+JkhIWPorM=; h=From:Subject:Date:References:To:In-Reply-To; b=p6Osm0NtVyea+DGRgzf7pzr2Mbd1kGbO7OM8MxAYgUGqFc8tBN3T514SLd4jVPIVY lI8LEy2haeInZC+EgydnXLSFdeXEiwqB8KUteN1AHc1Jm9Mi4Uj9qEqWGuUn7yPjwa OXLRXsC4pP7Ke77Cs4sU8lOBu8noBOjd65lZ+XB8=
From: Lars Eggert <lars@eggert.org>
Content-Type: multipart/signed; boundary="Apple-Mail=_40AE5EA3-EA21-4F04-BA8D-176BD117D5D1"; protocol="application/pgp-signature"; micalg=pgp-sha512
Mime-Version: 1.0 (Mac OS X Mail 13.4 \(3608.80.23.2.2\))
Subject: Wednesday, June 10 (was: Re: Doodle poll for -28 virtual interop event)
Date: Mon, 1 Jun 2020 13:47:07 +0300
References: <908284A4-51F5-42D5-B501-B4D81DA33AB0@eggert.org>
To: IETF QUIC WG <quic@ietf.org>
In-Reply-To: <908284A4-51F5-42D5-B501-B4D81DA33AB0@eggert.org>
Message-Id: <3FD0EDBA-1D00-4AB9-9DFA-768C864BC02A@eggert.org>
X-MailScanner-ID: EEA4C616E99.A4C3C
X-MailScanner: Found to be clean
X-MailScanner-From: lars@eggert.org
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/Opx9wAUkyfGAT4VyWNrOyeET6J0>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 01 Jun 2020 10:47:28 -0000

--Apple-Mail=_40AE5EA3-EA21-4F04-BA8D-176BD117D5D1
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=us-ascii

Hi,

On 2020-5-27, at 10:00, Lars Eggert <lars@eggert.org> wrote:
> please fill out https://doodle.com/poll/ubzzsne3ykppgse3 to find the =
best day(s) for a virtual interop on the -28 drafts.

Wednesday, June 10 got the broadest support. Mark your calendars.

Lars

--Apple-Mail=_40AE5EA3-EA21-4F04-BA8D-176BD117D5D1
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
	filename=signature.asc
Content-Type: application/pgp-signature;
	name=signature.asc
Content-Description: Message signed with OpenPGP

-----BEGIN PGP SIGNATURE-----
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=XLiP
-----END PGP SIGNATURE-----

--Apple-Mail=_40AE5EA3-EA21-4F04-BA8D-176BD117D5D1--


From nobody Mon Jun  1 17:14:28 2020
Return-Path: <lucaspardue.24.7@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 0A0DF3A16EA for <quic@ietfa.amsl.com>; Mon,  1 Jun 2020 17:14:26 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.848
X-Spam-Level: 
X-Spam-Status: No, score=-1.848 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_ENVFROM_END_DIGIT=0.25, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id NOsueo3Rgdvz for <quic@ietfa.amsl.com>; Mon,  1 Jun 2020 17:14:24 -0700 (PDT)
Received: from mail-wm1-x330.google.com (mail-wm1-x330.google.com [IPv6:2a00:1450:4864:20::330]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 399D33A16E9 for <quic@ietf.org>; Mon,  1 Jun 2020 17:14:24 -0700 (PDT)
Received: by mail-wm1-x330.google.com with SMTP id c71so1160246wmd.5 for <quic@ietf.org>; Mon, 01 Jun 2020 17:14:24 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:from:date:message-id:subject:to:cc; bh=3M4h8Z28hoGGN7lGwQ1oRS/8VTTPlpp64hOfSNL8PyY=; b=qjcrReqL1682nCYydZBqlKo0JAKZZqd77efCk8O6arFH3bbS4OzzjJu22n/xP4DEsK CTD26hirMWpxtcl3c5QAQLdY2x9AvUA6oi/k1PiV7vlrPF+mg/jkvUzcjAW0A1kLqAvZ dvn/uPLRpdJeDaIw3DsEg5RieShU+RwMDef07A7eZu4ATJls3N/q6+ML9WcrXcKvwxRc bXtBsTqHjkRHIpQ1pTHY744ErcnAN+SB9e3WPa7ApjyrnosJOtbTHsQ/L7dte/VpAsm6 vWtwx7jUytxI/1VRHa3hZ8NYrWlJnnEKKDeVSu/Jb2euoCU73yilRXBsJDR6dEEBEle0 aODw==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:from:date:message-id:subject:to:cc; bh=3M4h8Z28hoGGN7lGwQ1oRS/8VTTPlpp64hOfSNL8PyY=; b=cIuVfgjxUvzQjhmPrDZ4g0x3uVKravd3Zv+dn5q4gpoVI07JO8Y0MiVQNwj095ZRwe 6aivvwn/hKP5xFAaIHI12WeAhtDgNCP2mBQQw9TDFITtvGzsFn0ZvhPBA3nYfoMt9+kN elWGNXcXetckDHENAwxga49pTCsh0qLo5oT4CI7epLij8GjHQ/qRoMmQqAQLyi4fwzge xSWRt0vGiax+i1W9NMY3pVRsd8dNK4Cp5BeRG0pSQVzucndZypiK8GWo4d0dFYXLBhc+ 8KQAknDGWWoav0PSy5ZpIzNJHwQ7+l0DtoLJfN1w4zCdvtg2IuSbOjvSTXpfuMTJ0SVQ oTbA==
X-Gm-Message-State: AOAM531QEOlESc/zYUsAoCeTIYlTg7U6hhPMyZHwYsxB7F11iByeV/MV Ln1vahKx4WIg2BryqRqdSG0TVKE6BRrmOdMD8g3HaIpj
X-Google-Smtp-Source: ABdhPJzBrftAS/Fbm8BgvvwhbpVtx6+KlG+ihfmOsI0yXvlRvfnmrgCk0KzC5aLHzlP0VaBgkY7XwrQR/KDMBlcm0Iw=
X-Received: by 2002:a7b:c04f:: with SMTP id u15mr1559348wmc.98.1591056862384;  Mon, 01 Jun 2020 17:14:22 -0700 (PDT)
MIME-Version: 1.0
From: Lucas Pardue <lucaspardue.24.7@gmail.com>
Date: Tue, 2 Jun 2020 01:14:11 +0100
Message-ID: <CALGR9oZb7aTyFSFr-wg8-v5=rBgGPE0=RMj6Xed_HwPVoyoEOw@mail.gmail.com>
Subject: Consensus call for Late-stage documents, June 2020 edition
To: QUIC WG <quic@ietf.org>
Cc: Lars Eggert <lars@eggert.org>, Mark Nottingham <mnot@mnot.net>
Content-Type: multipart/alternative; boundary="0000000000001997ee05a70ecb8c"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/k-yFeZlYOJ9n-a2k_7fBkFbFvW4>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 02 Jun 2020 00:14:26 -0000

--0000000000001997ee05a70ecb8c
Content-Type: text/plain; charset="UTF-8"

The following issues have proposals for resolution, and discussion so far
seems
to support consensus to accept them. If you object, please do so on the
issue or
in response to this message (changing the Subject appropriately!).

The consensus call runs until close of business Eastern US time on Friday
June 5.

Absent any pushback, we'll direct the editors to action the proposals once
the
call concludes.

See <https://github.com/quicwg/base-drafts/projects/5> for the current
state of
issues in the Late Stage process, itself defined at
<https://github.com/quicwg/base-drafts/blob/master/CONTRIBUTING.md>.

* #3703: use TRANSPORT_PARAMETER_ERROR when authenticating connection IDs
  The proposal is https://github.com/quicwg/base-drafts/pull/3691

* #3698: Issue when a server cancels a push stream
  The proposal is https://github.com/quicwg/base-drafts/pull/3700

* #3695: Integrate QUIC DPLPMTUD Text
  The proposal is https://github.com/quicwg/base-drafts/pull/3702

* #3631: What error kind is caused by multiple GOAWAYS that increase the
limit
  The proposal is https://github.com/quicwg/base-drafts/pull/3634

* #3622: Make using static table and Huffman encoding in QPACK opt-in
  The proposal is to close this with no design change and mark the issue as
v2.

* #3619: Forgery limits on packet protection
  The proposal is https://github.com/quicwg/base-drafts/pull/3620

* #3608: 5-tuple routing and SPA
  The proposal is https://github.com/quicwg/base-drafts/pull/3670

* #3585: Text on ECN probing
  The proposal is to close with no design change. A separate editorial PR
  addresses a comment raised during discussion.

* #3529: Changing the Default QUIC ACK Policy
  The proposal is to close with no design change. An editorial change was
made
  to provide more description of the current design.

* #3526: QUIC PTO is too conservative, causing a measurable regression in
tail
latency
  The proposal is to close this with no design change and mark the issue as
v2.

* #3709: Make SERVER_BUSY more generic
  The proposal is to close with no action.

Regards Lars, Lucas and Mark
QUIC WG Chairs

--0000000000001997ee05a70ecb8c
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">The following issues have proposals for resolution, and di=
scussion so far seems<br>to support consensus to accept them. If you object=
, please do so on the issue or<br>in response to this message (changing the=
 Subject appropriately!). <br><br>The consensus call runs until close of bu=
siness Eastern US time on Friday June 5.<br><br>Absent any pushback, we&#39=
;ll direct the editors to action the proposals once the<br>call concludes.<=
br><br>See &lt;<a href=3D"https://github.com/quicwg/base-drafts/projects/5"=
>https://github.com/quicwg/base-drafts/projects/5</a>&gt; for the current s=
tate of<br>issues in the Late Stage process, itself defined at<br>&lt;<a hr=
ef=3D"https://github.com/quicwg/base-drafts/blob/master/CONTRIBUTING.md">ht=
tps://github.com/quicwg/base-drafts/blob/master/CONTRIBUTING.md</a>&gt;.<br=
><br>* #3703: use TRANSPORT_PARAMETER_ERROR when authenticating connection =
IDs<br>=C2=A0 The proposal is <a href=3D"https://github.com/quicwg/base-dra=
fts/pull/3691">https://github.com/quicwg/base-drafts/pull/3691<br></a><br>*=
 #3698: Issue when a server cancels a push stream<br>=C2=A0 The proposal is=
 <a href=3D"https://github.com/quicwg/base-drafts/pull/3700">https://github=
.com/quicwg/base-drafts/pull/3700</a><br><br>* #3695: Integrate QUIC DPLPMT=
UD Text<br>=C2=A0 The proposal is <a href=3D"https://github.com/quicwg/base=
-drafts/pull/3702">https://github.com/quicwg/base-drafts/pull/3702</a><br><=
br>* #3631: What error kind is caused by multiple GOAWAYS that increase the=
 limit<br>=C2=A0 The proposal is <a href=3D"https://github.com/quicwg/base-=
drafts/pull/3634">https://github.com/quicwg/base-drafts/pull/3634</a><br><b=
r>* #3622: Make using static table and Huffman encoding in QPACK opt-in<br>=
=C2=A0 The proposal is to close this with no design change and mark the iss=
ue as v2.<br><br>* #3619: Forgery limits on packet protection<br>=C2=A0 The=
 proposal is <a href=3D"https://github.com/quicwg/base-drafts/pull/3620">ht=
tps://github.com/quicwg/base-drafts/pull/3620</a><br><br>* #3608: 5-tuple r=
outing and SPA<br>=C2=A0 The proposal is <a href=3D"https://github.com/quic=
wg/base-drafts/pull/3670">https://github.com/quicwg/base-drafts/pull/3670</=
a><br><br>* #3585: Text on ECN probing<br>=C2=A0 The proposal is to close w=
ith no design change. A separate editorial PR<br>=C2=A0 addresses a comment=
 raised during discussion.<br><br>* #3529: Changing the Default QUIC ACK Po=
licy<br>=C2=A0 The proposal is to close with no design change. An editorial=
 change was made<br>=C2=A0 to provide more description of the current desig=
n. <br><br>* #3526: QUIC PTO is too conservative, causing a measurable regr=
ession in tail<br>latency<br>=C2=A0 The proposal is to close this with no d=
esign change and mark the issue as v2.<br><br>* #3709: Make SERVER_BUSY mor=
e generic<br>=C2=A0 The proposal is to close with no action.<br><br>Regards=
 Lars, Lucas and Mark<br> QUIC WG Chairs</div>

--0000000000001997ee05a70ecb8c--


From nobody Tue Jun  2 03:34:21 2020
Return-Path: <mirja.kuehlewind@ericsson.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id B969F3A00AD for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 03:34:18 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.001
X-Spam-Level: 
X-Spam-Status: No, score=-2.001 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, HTML_MESSAGE=0.001, HTTPS_HTTP_MISMATCH=0.1, RCVD_IN_MSPIKE_H2=-0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=ericsson.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id WmwMJZ0pDCF9 for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 03:34:15 -0700 (PDT)
Received: from EUR03-VE1-obe.outbound.protection.outlook.com (mail-eopbgr50057.outbound.protection.outlook.com [40.107.5.57]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 639EC3A003E for <quic@ietf.org>; Tue,  2 Jun 2020 03:34:15 -0700 (PDT)
ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=fdiMOauqHDOBbLueWkafSzMYjdmz+e5jUYM1Hj9qoN2x8lM3J4DiTOdAWcxoRSkClbHyRbd8AV7+LiZTXZo6dBLY63JkzoiC0y81p3Pd4KimgeNTclY1odLGPFutArgR5M0NZR62QxQmf8AoFT8ABBpcMUTmZELV/BOwMexhOswapGDhvl4XfcHpYa/LInsGmggoCApY4Fr/T2PBlo7RM0kH2LYLcOoM0KaAE0JIIQ8xVIu+lk8GYufEXXK4s4JXBR9gVdbmZZkZA+5hi7vX5goNsTdCV+5Ak/QVyUVTjaQ7MkT97JBxGv8JD2yM3jiwN8jb3aeOilqt8iOWE3zlcg==
ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com;  s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=ZzrhAC+hyBDin1+DbQ4bxN/Z+fn2RO8fwnsG/CDO/aE=; b=EYtBRkajP/ASe86ePhuSG+ZDFx83bvJQW8usHazS53mama0dG9uWwfyogMStR2Ytt5LvI7+QFDDpxwVwWfaoT5gHOxp1Rvx+T2pGUe3N2hISeW35kdZUhSENyE344HC84i/e4/0OLeLjNISAR/jecMxQ6SmzbrjifYcmkIi7KpOZxD2R56k86KqpgxzEoL7qaJGUS6IdBNH303jnZV5vc2ongJK/3563yh/uPMTGVDcNyn3YgfslNyyd1B12pa7fw8C3lxba704xw40ufUYmVlOHJe+RM4pzhwSzTO8ds87awLbWNvBSvrtrLKqUrOaZdMRC2KiBkf/4iEVxf5kbZg==
ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=ericsson.com; dmarc=pass action=none header.from=ericsson.com; dkim=pass header.d=ericsson.com; arc=none
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ericsson.com; s=selector1; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=ZzrhAC+hyBDin1+DbQ4bxN/Z+fn2RO8fwnsG/CDO/aE=; b=h0MkWsCmU4Ds/szeQwMKUr3+oJpFUpIBu9bg/SW3f2ufPHKKxbGlMrhaLrvkSGFrI8KzMFxcOxqG2VSXhZjS0bZpr5IKqMbFCsNuF1mp5TyexPlhkcHH7Rn/e3NozMFqZAGnH5XJwPjma2DgOWcHy7YGfdDbozmhcQaREdlGFGA=
Received: from AM0PR07MB4691.eurprd07.prod.outlook.com (2603:10a6:208:75::30) by AM0PR07MB4113.eurprd07.prod.outlook.com (2603:10a6:208:49::27) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.3066.12; Tue, 2 Jun 2020 10:34:10 +0000
Received: from AM0PR07MB4691.eurprd07.prod.outlook.com ([fe80::5c92:87f1:124a:e912]) by AM0PR07MB4691.eurprd07.prod.outlook.com ([fe80::5c92:87f1:124a:e912%5]) with mapi id 15.20.3066.016; Tue, 2 Jun 2020 10:34:10 +0000
From: Mirja Kuehlewind <mirja.kuehlewind@ericsson.com>
To: Nick Banks <nibanks=40microsoft.com@dmarc.ietf.org>, Martin Duke <martin.h.duke@gmail.com>, IETF QUIC WG <quic@ietf.org>
Subject: Re: QUIC-LB issues
Thread-Topic: QUIC-LB issues
Thread-Index: AQHWM4peMP6XQ3CTwkSqaQwuavGX3ai9lQcAgAe4HgA=
Date: Tue, 2 Jun 2020 10:34:10 +0000
Message-ID: <E1EB91DF-54F9-43E7-90F3-42ED042CB71D@ericsson.com>
References: <CAM4esxQG=YXThMA6uW3Y=3Qjt5xv8vOt92+FQ7GebcFGFVoH_g@mail.gmail.com> <SN2PR00MB0173119C2F79216A5C0D3750B38E0@SN2PR00MB0173.namprd00.prod.outlook.com>
In-Reply-To: <SN2PR00MB0173119C2F79216A5C0D3750B38E0@SN2PR00MB0173.namprd00.prod.outlook.com>
Accept-Language: en-US
Content-Language: en-GB
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
user-agent: Microsoft-MacOutlook/16.37.20051002
authentication-results: dmarc.ietf.org; dkim=none (message not signed) header.d=none;dmarc.ietf.org; dmarc=none action=none header.from=ericsson.com;
x-originating-ip: [2003:de:e700:7a00:4b4:966:e26f:f909]
x-ms-publictraffictype: Email
x-ms-office365-filtering-correlation-id: daa80ef6-2e34-4626-a73c-08d806e07c97
x-ms-traffictypediagnostic: AM0PR07MB4113:
x-microsoft-antispam-prvs: <AM0PR07MB41133F2765095DE73DD80B45F48B0@AM0PR07MB4113.eurprd07.prod.outlook.com>
x-ms-oob-tlc-oobclassifiers: OLM:10000;
x-forefront-prvs: 0422860ED4
x-ms-exchange-senderadcheck: 1
x-microsoft-antispam: BCL:0;
x-microsoft-antispam-message-info: TKNDQXmt6LkPWj4DEExkArf+a3p3zp/mUnDfHHwUivAlTfY6IDMY3MNOZ4YOuT4fy3ijn1hSNc90rNRCokCJrCQ3AWIlZFpbhpFFNoG65V58VajcaAB1p3XHJDu+wxuCBzqY5lgkCTCPZmAu93S+AtcOW6Me7qsw9jyb5q40dHm7Uihh5XjwGa0DNSNJiUujZIajtpHU8TgnGiEN0vSpoZdWq6lKWo35eyRhAq2loVQeLd9BxWG8loJfFUVq9GMG/ncZRixAH/MaSL4J8f9ZP695VRyoMIVIH4W/a6tmRTNDH62kCNe5g9kFw2HSd6F6qq3VOUwH9CPQ1KsFsSviEkwVTljeCaafP9xIXqApkuIwNQD38iWLwCgJJ/A65IEiqRDDuB0DLjwDaMUVEiqLaQ==
x-forefront-antispam-report: CIP:255.255.255.255; CTRY:; LANG:en; SCL:1; SRV:;  IPV:NLI; SFV:NSPM; H:AM0PR07MB4691.eurprd07.prod.outlook.com; PTR:; CAT:NONE;  SFTY:; SFS:(4636009)(346002)(136003)(366004)(376002)(39860400002)(396003)(86362001)(110136005)(316002)(36756003)(5660300002)(478600001)(8936002)(33656002)(166002)(64756008)(66476007)(66556008)(66446008)(83380400001)(966005)(6512007)(8676002)(6486002)(53546011)(83080400001)(71200400001)(2906002)(6506007)(186003)(2616005)(3480700007)(66946007)(44832011)(76116006); DIR:OUT; SFP:1101; 
x-ms-exchange-antispam-messagedata: GAn5It0Top+VPJ2I6OYQIC6nVe8kTr6XXCAPfsYxbOkBjy72bapLAUx0jahzbdFDeRfhaBSJo8trhVtBeKJOH/THfWa/JMgMPkiIDjS1hUNOjm5PMgMELfhlPwh9+Ounm19DA0P4G6OUnj0891lMmKz5b8kC3iQ8cKQqeD3lCf7fTtsbf3IF3TzXegga152gRmBmqj/8EDN7rPDXUb7s26llm++nKGmZ8VPCo7WtjyDz1/eDOyKDpPQps8gjxClbKsLPXJCyM9lekIIyu15YBGhYyQJ69xWx8SvaH93hxVylKbgDQxmvZL3QILVWkUs2x64cPbs2rZnZuysQAlDeW3Ri+uamjJxhErg0lKaVOy8s7+nFw/HOl1IQZxiWf3SIJ+I8yKTRvMBdrJJrSPglZs2beY92fGar+V9C0JELsid7z7+PX3rNhHJ3zZO5b8+VIMfCTRt+mYCJJL5wR7nzQRd38gFnpwe+IXrOzRQxNfDVbtZkHbVUFXQxeVLgle2cgm9W6NonrejfWZqSxE1Fah68pjmd5rMZjvUYMr6jpKU=
x-ms-exchange-transport-forked: True
Content-Type: multipart/alternative; boundary="_000_E1EB91DF54F943E790F342ED042CB71Dericssoncom_"
MIME-Version: 1.0
X-OriginatorOrg: ericsson.com
X-MS-Exchange-CrossTenant-Network-Message-Id: daa80ef6-2e34-4626-a73c-08d806e07c97
X-MS-Exchange-CrossTenant-originalarrivaltime: 02 Jun 2020 10:34:10.4290 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Hosted
X-MS-Exchange-CrossTenant-id: 92e84ceb-fbfd-47ab-be52-080c6b87953f
X-MS-Exchange-CrossTenant-mailboxtype: HOSTED
X-MS-Exchange-CrossTenant-userprincipalname: /LUEN1wb6N1V4DNyAWmfWxsh6MCYUamNyjbs46TJ2vdIW0H6/E39I8PvNmNraBEdMGahq1UT+L1VifqvzztCgiw0BtPYubSBM3nsY7XEn0I=
X-MS-Exchange-Transport-CrossTenantHeadersStamped: AM0PR07MB4113
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/vyh-LowBd2FdsBqvBuUo3NhjvyA>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 02 Jun 2020 10:34:19 -0000

--_000_E1EB91DF54F943E790F342ED042CB71Dericssoncom_
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
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==

--_000_E1EB91DF54F943E790F342ED042CB71Dericssoncom_
Content-Type: text/html; charset="utf-8"
Content-ID: <46A7918B26C3D1418D62D07E07C934FB@eurprd07.prod.outlook.com>
Content-Transfer-Encoding: base64
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--_000_E1EB91DF54F943E790F342ED042CB71Dericssoncom_--


From nobody Tue Jun  2 04:39:27 2020
Return-Path: <spencerdawkins.ietf@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id D36613A0860 for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 04:39:25 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Wi6rG61RlCuE for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 04:39:24 -0700 (PDT)
Received: from mail-lj1-x22a.google.com (mail-lj1-x22a.google.com [IPv6:2a00:1450:4864:20::22a]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id CE7183A085B for <quic@ietf.org>; Tue,  2 Jun 2020 04:39:23 -0700 (PDT)
Received: by mail-lj1-x22a.google.com with SMTP id a25so12179677ljp.3 for <quic@ietf.org>; Tue, 02 Jun 2020 04:39:23 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to;  bh=qzeNq/GY2N3lROpN8CrVg7lm1V25F1kvaScadoidq7g=; b=d+fvHP5OrVhsl8ocsnk/h1NJV3ojAxCxD28dSx8VE5Os+vLs7751eqjxzJxWuH0HPB 3o9HUJdASSfzWuzJ6BR46IK0PHmDoZBHLII1BmsTK0mXZRJP/BkyXRh18oNaDKn1hLac SW6h7G4qTDT/y0K4TUHvzChFC4Qu6ccmeYM+EgvURID2b3oELdxSRTSyrz2YZ6h4R+Ae u1XZTWpo7AajRQfhlnbGMy4FuT2D/1A/JPQ+fge5xI4Txc+RT+fSDwKRRI5FWHtLqDHa tpIbSkLE+Yt1YW6spLwLpcQma+n1gv/FSIWwnvEDFVSG3Z69MxGcN2WqPjXRMInDmpp6 SAfg==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to; bh=qzeNq/GY2N3lROpN8CrVg7lm1V25F1kvaScadoidq7g=; b=rnOgo5UFx0sGINdiuxltoMQXStp0oRjzM/bk0z84VAZUH0n/z/SIeh2xEdFSdspFxe 6ntDgtJs1AJJ+7sbKWfy1+oDOX0fqJwoCjHXGwaW2L2yS+7QGmkkU6gJt0GtOfgG2SIL /zfbO6Htpnlg9Y+ZnhCpgy1lyohr2mNhXO9gQ2/Wsy4dcSxVGiaYN0BhMmdwDOASuTta zhi1MGevvHbwGa9McIw4dAeBEQirR99zc+vH32sykKXcZTMC5p1NL/6UsL25zPcMlhy0 YYfA4Jwnz1o4oUkgN0pdRfsZy1k3Jt+5r4vf7Ftv3Nn8q9uyVYJY9g7wQak/0Lvc9efd on0Q==
X-Gm-Message-State: AOAM533DO5na5aeGakWoNQC9DN7JrgBJZo28uDz/z51hlRSEir3oiDrR jkQOm9wIjUiw6FcoqX6UgOzFW/ymXxC8pgI5HG7m5n3FlYk=
X-Google-Smtp-Source: ABdhPJz42FUKn4Qg4LqdzVcR1RyrIpCKfajc0GQktchn83cisIswyMS35Gflc0VoyYCr2xM2WAxL4xQwfM+laTZF0ds=
X-Received: by 2002:a05:651c:2007:: with SMTP id s7mr4977091ljo.418.1591097961737;  Tue, 02 Jun 2020 04:39:21 -0700 (PDT)
MIME-Version: 1.0
References: <159084638843.27466.7915766554130545967@ietfa.amsl.com>
In-Reply-To: <159084638843.27466.7915766554130545967@ietfa.amsl.com>
From: Spencer Dawkins at IETF <spencerdawkins.ietf@gmail.com>
Date: Tue, 2 Jun 2020 06:38:55 -0500
Message-ID: <CAKKJt-eHQtgjc-zuO7vrGZ1Q2c7=3hetOb0FyqnEmbTDu1Uwuw@mail.gmail.com>
Subject: Fwd: New Version Notification for draft-bonaventure-quic-atsss-overview-00.txt
To: IETF QUIC WG <quic@ietf.org>
Content-Type: multipart/alternative; boundary="000000000000cff5fe05a7185c41"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/HVSguVCVxMtps4E9SAqe35h9V04>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 02 Jun 2020 11:39:26 -0000

--000000000000cff5fe05a7185c41
Content-Type: text/plain; charset="UTF-8"

Dear QUIC,

Several of us submitted this draft (which I emphasize has no special status
in IETF, 3GPP, or between IETF and 3GPP).It's being used as one of the
inputs to SA2#139E, which is underway now (and that also doesn't give it
any special status).

We are, of course, interested in comments, especially if we've gotten
something wrong about QUIC, directions for MP-QUIC, or apparent gaps.

The QUIC chairs have told us this is an appropriate thing to discuss on the
QUIC mailing list - we did ask.

Best,

Spencer, but not just Spencer

---------- Forwarded message ---------
From: <internet-drafts@ietf.org>
Date: Sat, May 30, 2020 at 8:46 AM
Subject: New Version Notification for
draft-bonaventure-quic-atsss-overview-00.txt
To: SungHoon Seo <sh.seo@kt.com>, Olivier Bonaventure <
Olivier.Bonaventure@uclouvain.be>, Qing An <anqing.aq@alibaba-inc.com>,
Markus Amend <markus.amend@telekom.de>, Andreas Kassler <
andreas.kassler@kau.se>, Spencer Dawkins <spencerdawkins.ietf@gmail.com>,
Mohamed Boucadair <mohamed.boucadair@orange.com>, Quentin De Coninck <
quentin.deconinck@uclouvain.be>, Mirja Kuehlewind <
mirja.kuehlewind@ericsson.com>, Olivier Bonaventure <
olivier.bonaventure@uclouvain.be>, Nicolas Keukeleire <
nicolas.keukeleire@tessares.net>, Maxime Piraux <maxime.piraux@uclouvain.be>



A new version of I-D, draft-bonaventure-quic-atsss-overview-00.txt
has been successfully submitted by Olivier Bonaventure and posted to the
IETF repository.

Name:           draft-bonaventure-quic-atsss-overview
Revision:       00
Title:          3GPP Access Traffic Steering Switching and Splitting
(ATSSS) - Overview for IETF Participants
Document date:  2020-05-30
Group:          Individual Submission
Pages:          29
URL:
https://www.ietf.org/internet-drafts/draft-bonaventure-quic-atsss-overview-00.txt
Status:
https://datatracker.ietf.org/doc/draft-bonaventure-quic-atsss-overview/
Htmlized:
https://tools.ietf.org/html/draft-bonaventure-quic-atsss-overview-00
Htmlized:
https://datatracker.ietf.org/doc/html/draft-bonaventure-quic-atsss-overview


Abstract:
   This document briefly presents the Access Traffic Steering,
   Switching, and Splitting (ATSSS) service being specified within the
   3rd Generation Partnership Project (3GPP).  The ATSSS service
   provides network support for multihomed devices to select a path for
   transmission (steer), move traffic from one path to another (switch),
   or use multiple paths simultaneously (split).  TS 23.501 specifies an
   ATSSS architecture for TCP traffic.

   This document presents a snap-shot of the ongoing discussion in the
   3GPP to enable ATSSS for non-TCP traffic, based on the use of QUIC,
   and assesses to what extent IETF specifications can be used to meet
   the ATSSS design goals.  Apparent gaps are also documented.




Please note that it may take a couple of minutes from the time of submission
until the htmlized version and diff are available at tools.ietf.org.

The IETF Secretariat

--000000000000cff5fe05a7185c41
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Dear QUIC,=C2=A0<div><br></div><div>Several of us submitte=
d this draft (which I emphasize has no special status in IETF, 3GPP, or bet=
ween IETF and 3GPP).It&#39;s being used as one of the inputs to SA2#139E, w=
hich is underway now (and that also doesn&#39;t give it any special status)=
.=C2=A0</div><div><br></div><div>We are, of course, interested in comments,=
 especially if we&#39;ve gotten something wrong about QUIC, directions for =
MP-QUIC, or apparent gaps.</div><div><br></div><div>The QUIC chairs have to=
ld us this is an appropriate thing to discuss on the QUIC mailing list - we=
 did ask.=C2=A0</div><div><br></div><div>Best,</div><div><br></div><div>Spe=
ncer, but not just Spencer<br><br><div class=3D"gmail_quote"><div dir=3D"lt=
r" class=3D"gmail_attr">---------- Forwarded message ---------<br>From: <sp=
an dir=3D"auto">&lt;<a href=3D"mailto:internet-drafts@ietf.org" target=3D"_=
blank">internet-drafts@ietf.org</a>&gt;</span><br>Date: Sat, May 30, 2020 a=
t 8:46 AM<br>Subject: New Version Notification for draft-bonaventure-quic-a=
tsss-overview-00.txt<br>To: SungHoon Seo &lt;<a href=3D"mailto:sh.seo@kt.co=
m" target=3D"_blank">sh.seo@kt.com</a>&gt;, Olivier Bonaventure &lt;<a href=
=3D"mailto:Olivier.Bonaventure@uclouvain.be" target=3D"_blank">Olivier.Bona=
venture@uclouvain.be</a>&gt;, Qing An &lt;<a href=3D"mailto:anqing.aq@aliba=
ba-inc.com" target=3D"_blank">anqing.aq@alibaba-inc.com</a>&gt;, Markus Ame=
nd &lt;<a href=3D"mailto:markus.amend@telekom.de" target=3D"_blank">markus.=
amend@telekom.de</a>&gt;, Andreas Kassler &lt;<a href=3D"mailto:andreas.kas=
sler@kau.se" target=3D"_blank">andreas.kassler@kau.se</a>&gt;, Spencer Dawk=
ins &lt;<a href=3D"mailto:spencerdawkins.ietf@gmail.com" target=3D"_blank">=
spencerdawkins.ietf@gmail.com</a>&gt;, Mohamed Boucadair &lt;<a href=3D"mai=
lto:mohamed.boucadair@orange.com" target=3D"_blank">mohamed.boucadair@orang=
e.com</a>&gt;, Quentin De Coninck &lt;<a href=3D"mailto:quentin.deconinck@u=
clouvain.be" target=3D"_blank">quentin.deconinck@uclouvain.be</a>&gt;, Mirj=
a Kuehlewind &lt;<a href=3D"mailto:mirja.kuehlewind@ericsson.com" target=3D=
"_blank">mirja.kuehlewind@ericsson.com</a>&gt;, Olivier Bonaventure &lt;<a =
href=3D"mailto:olivier.bonaventure@uclouvain.be" target=3D"_blank">olivier.=
bonaventure@uclouvain.be</a>&gt;, Nicolas Keukeleire &lt;<a href=3D"mailto:=
nicolas.keukeleire@tessares.net" target=3D"_blank">nicolas.keukeleire@tessa=
res.net</a>&gt;, Maxime Piraux &lt;<a href=3D"mailto:maxime.piraux@uclouvai=
n.be" target=3D"_blank">maxime.piraux@uclouvain.be</a>&gt;<br></div><br><br=
><br>
A new version of I-D, draft-bonaventure-quic-atsss-overview-00.txt<br>
has been successfully submitted by Olivier Bonaventure and posted to the<br=
>
IETF repository.<br>
<br>
Name:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0draft-bonaventure-quic-atsss-=
overview<br>
Revision:=C2=A0 =C2=A0 =C2=A0 =C2=A000<br>
Title:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 3GPP Access Traffic Steering Switc=
hing and Splitting (ATSSS) - Overview for IETF Participants<br>
Document date:=C2=A0 2020-05-30<br>
Group:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 Individual Submission<br>
Pages:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 29<br>
URL:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 <a href=3D"https://www.ietf.o=
rg/internet-drafts/draft-bonaventure-quic-atsss-overview-00.txt" rel=3D"nor=
eferrer" target=3D"_blank">https://www.ietf.org/internet-drafts/draft-bonav=
enture-quic-atsss-overview-00.txt</a><br>
Status:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0<a href=3D"https://datatracker.iet=
f.org/doc/draft-bonaventure-quic-atsss-overview/" rel=3D"noreferrer" target=
=3D"_blank">https://datatracker.ietf.org/doc/draft-bonaventure-quic-atsss-o=
verview/</a><br>
Htmlized:=C2=A0 =C2=A0 =C2=A0 =C2=A0<a href=3D"https://tools.ietf.org/html/=
draft-bonaventure-quic-atsss-overview-00" rel=3D"noreferrer" target=3D"_bla=
nk">https://tools.ietf.org/html/draft-bonaventure-quic-atsss-overview-00</a=
><br>
Htmlized:=C2=A0 =C2=A0 =C2=A0 =C2=A0<a href=3D"https://datatracker.ietf.org=
/doc/html/draft-bonaventure-quic-atsss-overview" rel=3D"noreferrer" target=
=3D"_blank">https://datatracker.ietf.org/doc/html/draft-bonaventure-quic-at=
sss-overview</a><br>
<br>
<br>
Abstract:<br>
=C2=A0 =C2=A0This document briefly presents the Access Traffic Steering,<br=
>
=C2=A0 =C2=A0Switching, and Splitting (ATSSS) service being specified withi=
n the<br>
=C2=A0 =C2=A03rd Generation Partnership Project (3GPP).=C2=A0 The ATSSS ser=
vice<br>
=C2=A0 =C2=A0provides network support for multihomed devices to select a pa=
th for<br>
=C2=A0 =C2=A0transmission (steer), move traffic from one path to another (s=
witch),<br>
=C2=A0 =C2=A0or use multiple paths simultaneously (split).=C2=A0 TS 23.501 =
specifies an<br>
=C2=A0 =C2=A0ATSSS architecture for TCP traffic.<br>
<br>
=C2=A0 =C2=A0This document presents a snap-shot of the ongoing discussion i=
n the<br>
=C2=A0 =C2=A03GPP to enable ATSSS for non-TCP traffic, based on the use of =
QUIC,<br>
=C2=A0 =C2=A0and assesses to what extent IETF specifications can be used to=
 meet<br>
=C2=A0 =C2=A0the ATSSS design goals.=C2=A0 Apparent gaps are also documente=
d.<br>
<br>
<br>
<br>
<br>
Please note that it may take a couple of minutes from the time of submissio=
n<br>
until the htmlized version and diff are available at <a href=3D"http://tool=
s.ietf.org" rel=3D"noreferrer" target=3D"_blank">tools.ietf.org</a>.<br>
<br>
The IETF Secretariat<br>
<br>
<br>
</div></div></div>

--000000000000cff5fe05a7185c41--


From nobody Tue Jun  2 09:15:12 2020
Return-Path: <pravb@microsoft.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 2B0123A0C58 for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 09:15:10 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2
X-Spam-Level: 
X-Spam-Status: No, score=-2 tagged_above=-999 required=5 tests=[BAYES_00=-1.9,  DKIMWL_WL_HIGH=-0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, HTML_MESSAGE=0.001, HTTPS_HTTP_MISMATCH=0.1, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=microsoft.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 1Ige1Q3I5xx9 for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 09:15:07 -0700 (PDT)
Received: from NAM06-BL2-obe.outbound.protection.outlook.com (mail-eopbgr650106.outbound.protection.outlook.com [40.107.65.106]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 941A33A0C52 for <quic@ietf.org>; Tue,  2 Jun 2020 09:15:07 -0700 (PDT)
ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=L0H1UiLYZ6BobBMyxzq2r2vgN9Y92Jfd/Kb7vBwIu5Al3+kZxJGlWXP5QxhPu/0fnQZ/thRm5KXAjBnaUMGjsrh28HTBaZYAq2OB8QEiUG51SP5P7WNyh41TVOIs+zxy5JIcSCNjudqn6TV/joHVd3WuXtoNcJwcXYIaVVVNJTOVIzpmcT0+h7OdWmllf3HUQqOaG/+NGhCmvALGhQ+HZ4qeLDhYbSVI58batcb1leKd519DzgKZvHHnZwbogGY6zNJazMpuquELsHeu+jljkdDElzclOYrg7+DeFcHoupZsoqFBg/5aAXZrzxYLcKbHOkH07OwEHGESkOE/JDEzHA==
ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com;  s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=UHjkbRG8TdN8LIPGCZQGbLnOEpsAfutux8ZwLhcMAE8=; b=MkvaKpDEpPupSqC/yTm/eB6YQquhrYzkXqQIf5H7g9zkvgEJynX1CXC17G4LNt2tGKcF9bWQSdrsJ6YBZJYEbJuDO5JbwD1PNpI4piXezf2R89cfp+PSgq84QR4/Y+uEAlLT5/9YUL1GoCUbA1qmJsHoQmZB/bZJKmjClkXONNbtRSqyUk24b9h6epPkkbK9mpp73VN3L1r42bkxZwYUXwMT37HLlTHq4towK3vqkvMPEiicGKaqe8GpPExH4dppnLgLmfsor4G5mfSnu7+bo0GsVpyQeImYbn4+BxjF1kvI4GV15Ydkg7ExNQzPkdw2af183iNOxMUDsgJrorZfjQ==
ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=microsoft.com; dmarc=pass action=none header.from=microsoft.com; dkim=pass header.d=microsoft.com; arc=none
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=selector2; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=UHjkbRG8TdN8LIPGCZQGbLnOEpsAfutux8ZwLhcMAE8=; b=FhmC7NpP10UmPdK7U+9OPnKPSHspGDetfm9FRlX+prQBr4Vj0IAJtHMpfCTguLYW6NdwuR/r4idZNVxdMzCPp+UOZCnnw3GUO2KoowP+ngrpmHfS6uvKPvM8oxqOnMBIPmtOOJ4LR6dUknzDgdJXouRMwC6qKjAcEabVSTdG5MA=
Received: from CH2PR00MB0726.namprd00.prod.outlook.com (2603:10b6:610:ad::12) by CH2PR00MB0778.namprd00.prod.outlook.com (2603:10b6:610:6c::23) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.3097.0; Tue, 2 Jun 2020 16:15:04 +0000
Received: from CH2PR00MB0726.namprd00.prod.outlook.com ([fe80::447b:b279:69f8:4a73]) by CH2PR00MB0726.namprd00.prod.outlook.com ([fe80::447b:b279:69f8:4a73%4]) with mapi id 15.20.3096.000; Tue, 2 Jun 2020 16:15:04 +0000
From: Praveen Balasubramanian <pravb@microsoft.com>
To: Mirja Kuehlewind <mirja.kuehlewind=40ericsson.com@dmarc.ietf.org>, Nick Banks <nibanks=40microsoft.com@dmarc.ietf.org>, Martin Duke <martin.h.duke@gmail.com>, IETF QUIC WG <quic@ietf.org>
Subject: RE: QUIC-LB issues
Thread-Topic: QUIC-LB issues
Thread-Index: AQHWNP5etLQapo8XQ0OW+Lyn+dYvaajFKLYAgABemyA=
Date: Tue, 2 Jun 2020 16:15:04 +0000
Message-ID: <CH2PR00MB0726D9D56D3B42831D74A620B68B0@CH2PR00MB0726.namprd00.prod.outlook.com>
References: <CAM4esxQG=YXThMA6uW3Y=3Qjt5xv8vOt92+FQ7GebcFGFVoH_g@mail.gmail.com> <SN2PR00MB0173119C2F79216A5C0D3750B38E0@SN2PR00MB0173.namprd00.prod.outlook.com> <E1EB91DF-54F9-43E7-90F3-42ED042CB71D@ericsson.com>
In-Reply-To: <E1EB91DF-54F9-43E7-90F3-42ED042CB71D@ericsson.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
msip_labels: MSIP_Label_f42aa342-8706-4288-bd11-ebb85995028c_Enabled=true; MSIP_Label_f42aa342-8706-4288-bd11-ebb85995028c_SetDate=2020-06-02T16:15:02Z;  MSIP_Label_f42aa342-8706-4288-bd11-ebb85995028c_Method=Standard; MSIP_Label_f42aa342-8706-4288-bd11-ebb85995028c_Name=Internal; MSIP_Label_f42aa342-8706-4288-bd11-ebb85995028c_SiteId=72f988bf-86f1-41af-91ab-2d7cd011db47; MSIP_Label_f42aa342-8706-4288-bd11-ebb85995028c_ActionId=e9cfe472-08aa-4094-92eb-c849a1d89feb; MSIP_Label_f42aa342-8706-4288-bd11-ebb85995028c_ContentBits=0
authentication-results: dmarc.ietf.org; dkim=none (message not signed) header.d=none;dmarc.ietf.org; dmarc=none action=none header.from=microsoft.com;
x-originating-ip: [2001:4898:80e8:8:95d0:3f73:f682:d302]
x-ms-publictraffictype: Email
x-ms-office365-filtering-ht: Tenant
x-ms-office365-filtering-correlation-id: 6cd2f9a1-a5eb-4aef-52b9-08d807101c28
x-ms-traffictypediagnostic: CH2PR00MB0778:
x-microsoft-antispam-prvs: <CH2PR00MB07786D662118BE63037AB8F8B68B0@CH2PR00MB0778.namprd00.prod.outlook.com>
x-ms-oob-tlc-oobclassifiers: OLM:10000;
x-forefront-prvs: 0422860ED4
x-ms-exchange-senderadcheck: 1
x-microsoft-antispam: BCL:0;
x-microsoft-antispam-message-info: 2wVGQE9hRTkOMEIYnFg6wS4Wc62lmcWr4cniOIiEZky6v/uKZIj1/G2O4zc5HG7+ezcpezm0QwsL6bF6Wc+31v5QXKa0s9XECnNsHXJrVkK2YwtX2oqAvWqtKkTWe4rFK3tqDGG9Sywxy3io1//yrZ9xC33kRmUozY2xAb32zgD66R4f3nsWRlbqVc0MjWCITyl82y5jU8yTNU0TRMGNZrC2/Si3UTNxo+qkEAJKE+8yCXwPtKVil5O3bTbvSPhN8JcBFvMxvHirucPOQaO9HpsU4O8MJw+Xj8FvjBtTOa77W/tca9stolw9VLS6beJHsXAJuD0TCqg0SstY4IxDz0cZH39O+uJbV3JusXQzquS6PNzzmstj3D6MbN0q2qwx7TaZYX15PeWu5Lhp4kG3NQ==
x-forefront-antispam-report: CIP:255.255.255.255; CTRY:; LANG:en; SCL:1; SRV:;  IPV:NLI; SFV:NSPM; H:CH2PR00MB0726.namprd00.prod.outlook.com; PTR:; CAT:NONE;  SFTY:; SFS:(4636009)(376002)(39860400002)(136003)(396003)(346002)(366004)(66446008)(8990500004)(966005)(76116006)(5660300002)(66946007)(52536014)(66556008)(55016002)(9686003)(66476007)(6506007)(53546011)(82960400001)(110136005)(82950400001)(316002)(83080400001)(3480700007)(83380400001)(166002)(7696005)(8936002)(86362001)(186003)(478600001)(33656002)(2906002)(64756008)(71200400001)(10290500003)(8676002); DIR:OUT; SFP:1102; 
x-ms-exchange-antispam-messagedata: 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
x-ms-exchange-transport-forked: True
Content-Type: multipart/alternative; boundary="_000_CH2PR00MB0726D9D56D3B42831D74A620B68B0CH2PR00MB0726namp_"
MIME-Version: 1.0
X-OriginatorOrg: microsoft.com
X-MS-Exchange-CrossTenant-AuthAs: Internal
X-MS-Exchange-CrossTenant-AuthSource: CH2PR00MB0726.namprd00.prod.outlook.com
X-MS-Exchange-CrossTenant-Network-Message-Id: 6cd2f9a1-a5eb-4aef-52b9-08d807101c28
X-MS-Exchange-CrossTenant-originalarrivaltime: 02 Jun 2020 16:15:04.4030 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Hosted
X-MS-Exchange-CrossTenant-id: 72f988bf-86f1-41af-91ab-2d7cd011db47
X-MS-Exchange-CrossTenant-mailboxtype: HOSTED
X-MS-Exchange-CrossTenant-userprincipalname: 4r/b4fxpD1vP58eI9s9iAhBYTjfHwXdBNtI+r6Rp33wl5dKi5D7WKlvfiKrEVZF09z4FVjsfvVhMNILjgyoBAWhkpkM7W/ZqVP7ubBzHL8U=
X-MS-Exchange-Transport-CrossTenantHeadersStamped: CH2PR00MB0778
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/FL_mULTpdSDrYPq-eP9mSsTgphg>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 02 Jun 2020 16:15:10 -0000

--_000_CH2PR00MB0726D9D56D3B42831D74A620B68B0CH2PR00MB0726namp_
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64

Pj7igJxDb25uZWN0aW9uIElEcyBNVVNUIE5PVCBjb250YWluIGFueSBpbmZvcm1hdGlvbiB0aGF0
IGNhbiBiZSB1c2VkIGJ5IGFuIGV4dGVybmFsIG9ic2VydmVyICh0aGF0IGlzLCBvbmUgdGhhdCBk
b2VzIG5vdCBjb29wZXJhdGUgd2l0aCB0aGUgaXNzdWVyKSB0byBjb3JyZWxhdGUgdGhlbSB3aXRo
IG90aGVyIGNvbm5lY3Rpb24gSURzIGZvciB0aGUgc2FtZSBjb25uZWN0aW9uLuKAnQ0KVGhhdOKA
mXMgcXVpdGUgYW4gb25lcm91cyByZXF1aXJlbWVudCBhbmQgaXRzIHVuY2xlYXIgd2hhdCBpcyBt
ZWFudCBieSBjb3JyZWxhdGlvbi4gQW5kIHNpbmNlIGl0IGhhcyBub3RoaW5nIHRvIGRvIHdpdGgg
aW50ZXJvcCBJIGRvbuKAmXQgdW5kZXJzdGFuZCBob3cgdGhlIE1VU1QgTk9UIGNhbiBiZSB2ZXJp
ZmllZC4NCg0KRnJvbTogUVVJQyA8cXVpYy1ib3VuY2VzQGlldGYub3JnPiBPbiBCZWhhbGYgT2Yg
TWlyamEgS3VlaGxld2luZA0KU2VudDogVHVlc2RheSwgSnVuZSAyLCAyMDIwIDM6MzQgQU0NClRv
OiBOaWNrIEJhbmtzIDxuaWJhbmtzPTQwbWljcm9zb2Z0LmNvbUBkbWFyYy5pZXRmLm9yZz47IE1h
cnRpbiBEdWtlIDxtYXJ0aW4uaC5kdWtlQGdtYWlsLmNvbT47IElFVEYgUVVJQyBXRyA8cXVpY0Bp
ZXRmLm9yZz4NClN1YmplY3Q6IFtFWFRFUk5BTF0gUmU6IFFVSUMtTEIgaXNzdWVzDQoNCkhpIE1h
cnRpbiwNCg0KSSBjYW7igJl0IHByb3ZpZGUgeW91IGFueSBmZWVkYmFjayBvbiBpbXBsZW1lbnRh
dGlvbiBvciBkZXBsb3ltZW50IHBsYW5zIGJ1dCBnaXZlbiB0aGVzZSBhbGdvcml0aG1zIGFyZSB1
c2VkIHdpdGggb25lIGRvbWFpbiwgSSB3b3VsZCBhc3N1bWUgdGhhdCBzb21lIHBsYWludGV4dCBh
bGdvcml0aG1zIHdvdWxkIGJlIHVzZWQgYW55d2F5IGV2ZW4gaWYgbm90IGRvY3VtZW50ZWQgaW4g
dGhpcyBkcmFmdC4gVGhlcmVmb3JlIEkgd291bGQgcmF0aGVyIGxpa2Ugc2VlIGFsbCB0aGUgZGlm
ZmVyZW50IGFwcHJvYWNoZXMgZG9jdW1lbnRlZCBhbmQgZGlzY3Vzc2VkIGhlcmUuIEFuZCB0aGVy
ZSBpcyBhbHJlYWR5IGEgY2xlYXIgd2FybmluZyBpbiB0aGUgc2VjdXJpdHkgY29uc2lkZXJhdGlv
biBzZWN0aW9uLg0KDQpOb3Qgc3VyZSBhYm91dCB0aGUgbmVlZCBmb3IgYSBuZXcgdHJhbnNwb3J0
IHBhcmFtZXRlci4gSSBndWVzcyBnaXZlbiB0aGUgc2VydmVyIHNlbGVjdHMgdGhlIENJRCB0aGVy
ZSBjb3VsZCBhbHdheXMgYmUgYSByaXNrIHRoYXQgdGhleSBhcmUgc29tZWhvdyBsaW5rYWJsZS4g
VGhlcmUgaXMgbm8gd2F5IGZvciB0aGUgY2xpZW50IHRvIGNoZWNrIHRoYXQuIFNvIGlmIGEgY2xp
ZW50IHJlYWxseSwgcmVhbGx5IHdhbnRzIHRvIGVuc3VyZSB0aGF0IHRoZXJlIGlzIG5vIGxpbmth
YmlsaXR5LCBpdCBwcm9iYWJseSBnZW5lcmFsbHkgaGFzIHRvIGRpc2FibGUgbWlncmF0aW9uIGVu
dGlyZWx5Lg0KDQpIb3dldmVyLCBJIHdvdWxkIGFsc28gbGlrZSB0byBub3RlIHRoYXQgdGhlIGN1
cnJlbnQgdHJhbnNwb3J0IGRyYWZ0IHNheXMgdGhlIGZvbGxvd2luZzoNCuKAnENvbm5lY3Rpb24g
SURzIE1VU1QgTk9UIGNvbnRhaW4gYW55IGluZm9ybWF0aW9uIHRoYXQgY2FuIGJlIHVzZWQgYnkg
YW4gZXh0ZXJuYWwgb2JzZXJ2ZXIgKHRoYXQgaXMsIG9uZSB0aGF0IGRvZXMgbm90IGNvb3BlcmF0
ZSB3aXRoIHRoZSBpc3N1ZXIpIHRvIGNvcnJlbGF0ZSB0aGVtIHdpdGggb3RoZXIgY29ubmVjdGlv
biBJRHMgZm9yIHRoZSBzYW1lIGNvbm5lY3Rpb24u4oCdDQoNCk1pcmphDQoNCg0KDQpGcm9tOiBR
VUlDIDxxdWljLWJvdW5jZXNAaWV0Zi5vcmc8bWFpbHRvOnF1aWMtYm91bmNlc0BpZXRmLm9yZz4+
IG9uIGJlaGFsZiBvZiBOaWNrIEJhbmtzIDxuaWJhbmtzPTQwbWljcm9zb2Z0LmNvbUBkbWFyYy5p
ZXRmLm9yZzxtYWlsdG86bmliYW5rcz00MG1pY3Jvc29mdC5jb21AZG1hcmMuaWV0Zi5vcmc+Pg0K
RGF0ZTogVGh1cnNkYXksIDI4LiBNYXkgMjAyMCBhdCAxNjo0MQ0KVG86IE1hcnRpbiBEdWtlIDxt
YXJ0aW4uaC5kdWtlQGdtYWlsLmNvbTxtYWlsdG86bWFydGluLmguZHVrZUBnbWFpbC5jb20+Piwg
SUVURiBRVUlDIFdHIDxxdWljQGlldGYub3JnPG1haWx0bzpxdWljQGlldGYub3JnPj4NClN1Ympl
Y3Q6IFJFOiBRVUlDLUxCIGlzc3Vlcw0KDQpIZWxsbyBNYXJ0aW4gJiBXRywNCg0KRm9yIEF6dXJl
LCB3ZeKAmXJlIGluIHRoZSBwcm9jZXNzIG9mIGJ1aWxkaW5nIGEgbG9hZCBiYWxhbmNpbmcgc2No
ZW1lIG9uIHRvcCBvZiB0aGUgUGxhaW50ZXh0IENJRCBhbGdvcml0aG0sIHNvIHdl4oCZZCBkZWZp
bml0ZWx5IGxpa2UgdG8ga2VlcCBpdCBpbiB0aGUgZG9jdW1lbnQuIFRoZSBjb21wbGljYXRlZCBu
YXR1cmUgb2Ygc29tZSBvZiBvdXIgc2NlbmFyaW9zIChtdWx0aXBsZSBjb250YWluZXJzIHBlci1E
SVAsIGFsbCBiZWhpbmQgYSBzaW5nbGUgTEI7IHdpdGggYWxsIGNvbXBvbmVudHMgZ2VuZXJhbGx5
IGluZGVwZW5kZW50KSBjdXJyZW50bHkgcHJldmVudHMgdXMgZnJvbSBwcmFjdGljYWxseSBzdXBw
b3J0aW5nIGFueXRoaW5nIG1vcmUgdGhhbiB0aGUgbW9zdCBiYXNpYyBhbGdvcml0aG0uIEFkZGl0
aW9uYWxseSwgd2hldGhlciBpdOKAmXMgYWRvcHRlZCBieSB0aGUgV0cgb3Igc29tZXRoaW5nIEF6
dXJlIHNwZWNpZmljLCB3ZSB3aWxsIGxpa2VseSByZXF1aXJlIGFuIGluLWJhbmQgY29tbXVuaWNh
dGlvbiB0byBjb21tdW5pY2F0ZSB0aGUgU2VydmVyIElEIHRvIGFsbCB0aGUgcm91dGluZyBjb21w
b25lbnRzIGludm9sdmVkLg0KDQpXZSBtYXkgZXZlbnR1YWxseSBhZG9wdCBvbmUgb2YgdGhlIGNp
cGhlcnRleHQgYWxnb3JpdGhtcyBpbiB0aGUgZnV0dXJlLCBidXQgSSBkb3VidCB3ZeKAmWQgZXZl
ciBhZGQgc3VwcG9ydCBmb3IgdGhlIE9iZnVzY2F0ZWQgQ0lEIGFsZ29yaXRobS4NCg0KT24gdGhl
IHRvcGljIG9mIGlzc3VlICM4LCB3ZSBkaXNhZ3JlZSB0aGF0IHRoZSBwb3NzaWJsZSBpbmZvcm1h
dGlvbiBleHBvc3VyZSB3aGVuIHVzaW5nIHRoZSBQbGFpbnRleHQgQ0lEIGFsZ29yaXRobSBpcyBl
bm91Z2ggcmVhc29uIHRvIGRlY2xhcmUgdGhlIENJRHMgYXMg4oCcZ3Vlc3NhYmxl4oCdLiBTbyB3
ZeKAmWQgYmUgYWdhaW5zdCByZWNvbW1lbmRpbmcvcmVxdWlyaW5nIHRoYXQgbWlncmF0aW9uIGJl
IGRpc2FibGVkIHdoZW4gdXNpbmcgdGhlIHBsYWludGV4dCBhbGdvcml0aG0uIEFkZGl0aW9uYWxs
eSwgaW4gb3VyIG9waW5pb24sIGlmIG1pZ3JhdGlvbiBpcyBub3Qgc3VwcG9ydGVkL2FsbG93ZWQg
YnkgdGhlIHNlcnZlciwgdGhlbiB0aGVyZSBpc27igJl0IHJlYWxseSBhbnkgcmVhc29uIHRvIGRv
IGFueXRoaW5nIG1vcmUgdGhhbiBJUCB0dXBsZSBsb2FkIGJhbGFuY2luZyBhdCBhbGwuIFRoZSB3
aG9sZSBwb2ludCBvZiBRVUlDLUxCIGlzIHRvIHByb3ZpZGUgZm9yIGxvYWQgYmFsYW5jaW5nIHNv
bHV0aW9ucyB0aGF0IHN1cnZpdmUgdHVwbGUgY2hhbmdlcy4NCg0KVGhhbmtzLA0KLSBOaWNrDQoN
CkZyb206IFFVSUMgPHF1aWMtYm91bmNlc0BpZXRmLm9yZzxtYWlsdG86cXVpYy1ib3VuY2VzQGll
dGYub3JnPj4gT24gQmVoYWxmIE9mIE1hcnRpbiBEdWtlDQpTZW50OiBUdWVzZGF5LCBNYXkgMjYs
IDIwMjAgMTE6MjAgQU0NClRvOiBJRVRGIFFVSUMgV0cgPHF1aWNAaWV0Zi5vcmc8bWFpbHRvOnF1
aWNAaWV0Zi5vcmc+Pg0KU3ViamVjdDogUVVJQy1MQiBpc3N1ZXMNCg0KSGVsbG8gYWxsLA0KDQpU
aGVyZSBoYXNuJ3QgYmVlbiBtdWNoIGFjdGl2aXR5IGluIFFVSUMtTEI8aHR0cHM6Ly9uYW0wNi5z
YWZlbGlua3MucHJvdGVjdGlvbi5vdXRsb29rLmNvbS8/dXJsPWh0dHBzJTNBJTJGJTJGcXVpY3dn
LmdpdGh1Yi5pbyUyRmxvYWQtYmFsYW5jZXJzJTJGZHJhZnQtaWV0Zi1xdWljLWxvYWQtYmFsYW5j
ZXJzLmh0bWwmZGF0YT0wMiU3QzAxJTdDcHJhdmIlNDBtaWNyb3NvZnQuY29tJTdDZjBhOTUxZTcw
ZjY2NGMzYTc2MGMwOGQ4MDZlMDg5YmElN0M3MmY5ODhiZjg2ZjE0MWFmOTFhYjJkN2NkMDExZGI0
NyU3QzElN0MwJTdDNjM3MjY2OTA4NzU2NjM2Mzg0JnNkYXRhPWNOOW1RNHZGQ0g3aDhEMVRPVkFt
ZFN6ZVgwRUZTJTJGWiUyRmhkZjRSTmVLSnc0JTNEJnJlc2VydmVkPTA+IGxhdGVseS4gVGhlcmUg
aXMgYXQgbGVhc3Qgb25lIGlzc3VlIHRoYXQgY291bGQgdXNlIHNvbWUgYnJvYWRlciBjb21tdW5p
dHkgaW5wdXQuDQoNCmh0dHBzOi8vZ2l0aHViLmNvbS9xdWljd2cvbG9hZC1iYWxhbmNlcnMvaXNz
dWVzLzg8aHR0cHM6Ly9uYW0wNi5zYWZlbGlua3MucHJvdGVjdGlvbi5vdXRsb29rLmNvbS8/dXJs
PWh0dHBzJTNBJTJGJTJGZ2l0aHViLmNvbSUyRnF1aWN3ZyUyRmxvYWQtYmFsYW5jZXJzJTJGaXNz
dWVzJTJGOCZkYXRhPTAyJTdDMDElN0NwcmF2YiU0MG1pY3Jvc29mdC5jb20lN0NmMGE5NTFlNzBm
NjY0YzNhNzYwYzA4ZDgwNmUwODliYSU3QzcyZjk4OGJmODZmMTQxYWY5MWFiMmQ3Y2QwMTFkYjQ3
JTdDMSU3QzAlN0M2MzcyNjY5MDg3NTY2NDYzNzcmc2RhdGE9JTJCUyUyRmJjQ2NGbEo4eXNUbXhm
V05CRFRGWEkwemhabmdka0dFVVdndGtKb1ElM0QmcmVzZXJ2ZWQ9MD4NCg0KVGhlIGNvcmUgaXNz
dWUgaXMgdGhhdCB0aGVyZSBhcmUgNCBhbGdvcml0aG1zIHRvIGVuY29kZSB0aGUgc2VydmVyIElE
LCAyIGNpcGhlcnRleHQgYW5kIDIgcGxhaW50ZXh0LiBPZiB0aGUgdHdvIHBsYWludGV4dCBhbGdv
cml0aG1zLCBvbmUgaXMgYmxpbmRpbmdseSBvYnZpb3VzIChqdXN0IGVuY29kZSBpdCBhIGtub3du
IGZpZWxkKSBhbmQgdGhlIG90aGVyIGFwcGxpZXMgcXVpdGUgYSBiaXQgb2Ygb2JmdXNjYXRpb24u
IFRoZSBvYmZ1c2NhdGVkIHZlcnNpb24gaGFzIHVuZGVyZ29uZSBzb21lIGluZm9ybWFsIGFuYWx5
c2lzIHZzLiBicnV0ZSBmb3JjZSBhdHRhY2tzLCBidXQgSSBkb3VidCBjcnlwdG8gZXhwZXJ0cyB3
b3VsZCBiZSBpbXByZXNzZWQuDQoNClRoZSBRVUlDIFdHIHJlZmxleCBoYXMgZ2VuZXJhbGx5IGJl
ZW4gdG8gZW5jcnlwdCBldmVyeXRoaW5nIHBvc3NpYmxlLiBUaGVyZSBhcmUgdHdvIHJlYXNvbnMg
dGhlc2VwbGFpbnRleHQgYWxnb3JpdGhtcyBleGlzdDoNCjEpIEFzIHNlY3Rpb24gMi4yIGRlc2Ny
aWJlczxodHRwczovL25hbTA2LnNhZmVsaW5rcy5wcm90ZWN0aW9uLm91dGxvb2suY29tLz91cmw9
aHR0cHMlM0ElMkYlMkZxdWljd2cub3JnJTJGbG9hZC1iYWxhbmNlcnMlMkZkcmFmdC1pZXRmLXF1
aWMtbG9hZC1iYWxhbmNlcnMuaHRtbCUyM25hbWUtc2VjdXJpdHkmZGF0YT0wMiU3QzAxJTdDcHJh
dmIlNDBtaWNyb3NvZnQuY29tJTdDZjBhOTUxZTcwZjY2NGMzYTc2MGMwOGQ4MDZlMDg5YmElN0M3
MmY5ODhiZjg2ZjE0MWFmOTFhYjJkN2NkMDExZGI0NyU3QzElN0MwJTdDNjM3MjY2OTA4NzU2NjQ2
Mzc3JnNkYXRhPUhqRVVMNEdzYzVtYklQQkZWSGk4Q1g5ZE1seVRwMVl5TDVmSjF1YUtLaTAlM0Qm
cmVzZXJ2ZWQ9MD4sIHRoZXJlIGlzIG5vIHN1Y2ggdGhpbmcgYXMgcGVyZmVjdCBwcm90ZWN0aW9u
IGZyb20gbGlua2FiaWxpdHksIHNvIGV2ZW4gZnVsbCBlbmNyeXB0aW9uIGlzIG9ubHkgbW92aW5n
IHRoaW5ncyBhbG9uZyBhIGNvbnRpbnV1bS4gSXQgaXMgbm90IG5lYXJseSBhcyBicmlnaHQgYSBs
aW5lIGFzLCBzYXksIGh0dHBzIHZzLiBodHRwLg0KMikgV2hhdCBsaXR0bGUgZmVlZGJhY2sgd2Un
dmUgZ290dGVuIGZyb20gcGVvcGxlIHRoYXQgbWFrZSBsYXllciA0IGNsb3VkIGxvYWQgYmFsYW5j
ZXJzIGhhcyBiZWVuIHByZXR0eSBuZWdhdGl2ZSBhYm91dCBhZGRpbmcgZW5jcnlwdGlvbiB0byB0
aGVpciBpbXBsZW1lbnRhdGlvbnMuIEkgY2FuIHByb2JhYmx5IGdldCBGNSB0byBhZG9wdCBpdCwg
YnV0IG90aGVyIHRoYW4gdGhhdCB0aGUgZGVwbG95bWVudCBzdG9yeSBpcyBub3Qgc29saWQgb24g
dGhlIGxvYWQgYmFsYW5jZXIgc2lkZS4gVGhlcmUgaXMgc3RpbGwgbXVjaCB3b3JrIHRvIGRvIGhl
cmUgb24gb3V0cmVhY2ggdG8gdGhlIGJpZyBjbG91ZCBwcm92aWRlcnMuDQoNClBvc3NpYmx5LCBh
biBvcGVuIHNvdXJjZSBpbXBsZW1lbnRhdGlvbiAoSSdtIHdvcmtpbmcgb24gaXQuLi4pIG1pZ2h0
IGFzc3VhZ2UgdGhlaXIgZmVhcnMuIEFuIGltcGxlbWVudGF0aW9uIHRoYXQgaW1wbGVtZW50cyBi
b3RoIExCLXNpZGUgZW5jcnlwdGVkIGFsZ29yaXRobXMgaXMgbGVzcyB0aGFuIDgwIGxpbmVzIG9m
IEMsIG5vdCBpbmNsdWRpbmcgb3BlbnNzbCBvciB0aGUgc2VwYXJhdGUgY29kZSB0byBjb25maWd1
cmUgdGhlIExCIHByb3Blcmx5Lg0KDQpUaGUgZGVzaWduIHRlYW0gaGFkIGNvbnNlbnN1cyB0byBp
bmNsdWRlIGFsbCBvZiB0aGVzZSBvcHRpb25zLCB3aXRoIHNvbWUgaW50ZW5kaW5nIHRvIHN1cHBv
cnQgdGhlbSBpbiB0aGVpciBwcm9kdWN0cy4gQmV5b25kIHRoYXQsIE1hcnRpbiBUIGlzIHRoZSBv
bmx5IHBlcnNvbiBJJ20gYXdhcmUgb2YgdG8gaGF2ZSBlbmdhZ2VkIG9uIHRoZSBQVCBvcHRpb25z
LCBhbmQgaGUgd2FzIHByZXR0eSBuZWdhdGl2ZSBhYm91dCB0aGVtLiBXaGF0IGlzIHRoZSBzZW5z
ZSBvZiB0aGUgZ3JvdXAgYXMgYSB3aG9sZT8NCjEuIFNob3VsZCB0aGlzIGRvY3VtZW50IGluY2x1
ZGUgdGhlIFBsYWludGV4dCBDSUQgKFBDSUQpIGFsZ29yaXRobTxodHRwczovL25hbTA2LnNhZmVs
aW5rcy5wcm90ZWN0aW9uLm91dGxvb2suY29tLz91cmw9aHR0cHMlM0ElMkYlMkZxdWljd2cub3Jn
JTJGbG9hZC1iYWxhbmNlcnMlMkZkcmFmdC1pZXRmLXF1aWMtbG9hZC1iYWxhbmNlcnMuaHRtbCUy
M25hbWUtcGxhaW50ZXh0LWNpZC1hbGdvcml0aG0mZGF0YT0wMiU3QzAxJTdDcHJhdmIlNDBtaWNy
b3NvZnQuY29tJTdDZjBhOTUxZTcwZjY2NGMzYTc2MGMwOGQ4MDZlMDg5YmElN0M3MmY5ODhiZjg2
ZjE0MWFmOTFhYjJkN2NkMDExZGI0NyU3QzElN0MwJTdDNjM3MjY2OTA4NzU2NjU2MzcyJnNkYXRh
PSUyRlBVcUZyRGthM0J1Nk10dHVlbWNCRHRQa1kzOFpMU2dXYkFRTWUlMkI1emlBJTNEJnJlc2Vy
dmVkPTA+Pw0KMi4gU2hvdWxkIHRoaXMgZG9jdW1lbnQgaW5jbHVkZSB0aGUgT2JmdXNjYXRlZCBD
SUQgKE9DSUQpIGFsZ29yaXRobTxodHRwczovL25hbTA2LnNhZmVsaW5rcy5wcm90ZWN0aW9uLm91
dGxvb2suY29tLz91cmw9aHR0cHMlM0ElMkYlMkZxdWljd2cub3JnJTJGbG9hZC1iYWxhbmNlcnMl
MkZkcmFmdC1pZXRmLXF1aWMtbG9hZC1iYWxhbmNlcnMuaHRtbCUyM25hbWUtb2JmdXNjYXRlZC1j
aWQtYWxnb3JpdGhtJmRhdGE9MDIlN0MwMSU3Q3ByYXZiJTQwbWljcm9zb2Z0LmNvbSU3Q2YwYTk1
MWU3MGY2NjRjM2E3NjBjMDhkODA2ZTA4OWJhJTdDNzJmOTg4YmY4NmYxNDFhZjkxYWIyZDdjZDAx
MWRiNDclN0MxJTdDMCU3QzYzNzI2NjkwODc1NjY1NjM3MiZzZGF0YT1ndlppRGVZb2poaVJGTW1j
amxaWlY0R2w1S2V6cDZBeURIbTF5RFpTSHJVJTNEJnJlc2VydmVkPTA+Pw0KDQoqKioqKg0KDQpP
biBhIHJlbGF0ZWQgbm90ZTogb25lIGlzc3VlIHdpdGggdGhlIHBsYWludGV4dC9vYmZ1c2NhdGVk
IGFsZ29yaXRobXMgaXMgdGhhdCB0aGV5IGFyZSBjaG9zZW4gYnkgdGhlIHNlcnZlciBidXQgdGhl
IGNvbnNlcXVlbmNlcyBvZiBsaW5rYWJpbGl0eSBhcmUgYm9ybmUgYnkgdGhlIGNsaWVudC4gVGhl
IGN1cnJlbnQgZHJhZnQgc2F5cyB0aGF0IHRoZSBub24tb2JmdXNjYXRlZCBhbGdvcml0aG0gU0hP
VUxEIGJlIGFjY29tcGFuaWVkIGJ5IGEgZGlzYWJsZV9hY3RpdmVfbWlncmF0aW9uIHBhcmFtZXRl
cjsgdGhhdCdzIHBvdGVudGlhbGx5IGFuIG92ZXJsb2FkZWQgY29kZXBvaW50Lg0KDQpodHRwczov
L2dpdGh1Yi5jb20vcXVpY3dnL2xvYWQtYmFsYW5jZXJzL2lzc3Vlcy8xNjxodHRwczovL25hbTA2
LnNhZmVsaW5rcy5wcm90ZWN0aW9uLm91dGxvb2suY29tLz91cmw9aHR0cHMlM0ElMkYlMkZnaXRo
dWIuY29tJTJGcXVpY3dnJTJGbG9hZC1iYWxhbmNlcnMlMkZpc3N1ZXMlMkYxNiZkYXRhPTAyJTdD
MDElN0NwcmF2YiU0MG1pY3Jvc29mdC5jb20lN0NmMGE5NTFlNzBmNjY0YzNhNzYwYzA4ZDgwNmUw
ODliYSU3QzcyZjk4OGJmODZmMTQxYWY5MWFiMmQ3Y2QwMTFkYjQ3JTdDMSU3QzAlN0M2MzcyNjY5
MDg3NTY2NjYzNzYmc2RhdGE9TjIlMkI0c05xSnpCekVJRiUyRmdsNUhjMXA2N0R3WUUlMkJQZG0w
Sk1GOGI2Tmt4NCUzRCZyZXNlcnZlZD0wPg0KDQpBbiBhbHRlcm5hdGl2ZSB3b3VsZCBiZSB0byBj
cmVhdGUgYSBuZXcgdHJhbnNwb3J0IHBhcmFtZXRlciB0aGF0IHdvdWxkIGFsbG93IHRoZSBzZXJ2
ZXIgdG8gZXhwbGljaXRseSBzYXkgd2hhdCBpdCBpcyBkb2luZy4gQ2xpZW50cyBjb3VsZCB0aGVu
IGRlY2lkZSB3aGV0aGVyIG9yIG5vdCB0byBtaWdyYXRlIHdpdGggdGhlaXIgZXllcyB3aWRlIG9w
ZW4uIE9uIHRoZSBvdGhlciBoYW5kLCB0aGlzIGxlYWtzIHNvbWUgaW5mb3JtYXRpb24gYWJvdXQg
dGhlIGVuY29kaW5nIGludG8gdGhlIG5ldHdvcmsuDQoNCjMuIEFzc3VtaW5nIHdlIGhhdmUgYXQg
bGVhc3Qgb25lIHVuZW5jcnlwdGVkIGFsZ29yaXRobSwgaXMgdGhpcyBzZXJ2ZXIgVFAgYSBnb29k
IGlkZWEgb3Igbm90Pw0K

--_000_CH2PR00MB0726D9D56D3B42831D74A620B68B0CH2PR00MB0726namp_
Content-Type: text/html; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_CH2PR00MB0726D9D56D3B42831D74A620B68B0CH2PR00MB0726namp_--


From nobody Tue Jun  2 09:34:58 2020
Return-Path: <matt.joras@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 26FBF3A0D4D for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 09:34:50 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id rprT9Cvp5mN7 for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 09:34:47 -0700 (PDT)
Received: from mail-vk1-xa2d.google.com (mail-vk1-xa2d.google.com [IPv6:2607:f8b0:4864:20::a2d]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id D07293A0DC3 for <quic@ietf.org>; Tue,  2 Jun 2020 09:34:27 -0700 (PDT)
Received: by mail-vk1-xa2d.google.com with SMTP id e1so1057273vkd.1 for <quic@ietf.org>; Tue, 02 Jun 2020 09:34:27 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=TjKRzpcasSMBURlt8qxicwvNa7YN1BqFr6H14fr2dl8=; b=Q24EXJbd1WhXmk61Z1/WH4vebhIZ9AQNx0AQUKoVTFoeVZbj/5Vc8Gn9euivgvfrnB NLhx5KD/5lZsV6OEQQiLk6zke6pEK9EWQGC1JO35EwjJ4u1xUqwHi7Lyf2hJV0YV+IOd d/0mlp8aVtwwIRga8R9n6BksfiftEkeLC0YS/k/WkTuZ34k21+ZLsrb6DCxA/XMqc1Lq PEBzWRzxAUF9bUGqdci2Xw9UubsMwmRuXo2m8j07V2hOHozRYiLAhyb168SOXw3IZeX7 BMFB7W6a1P13W1akyvXyu0YfZco8uuHfYC5LlqY1CDGAZ7Eo+2cqFMUoS4Nr9ad/dcc8 QbMw==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=TjKRzpcasSMBURlt8qxicwvNa7YN1BqFr6H14fr2dl8=; b=U4OyMv95fPIz4mHaOSuryX87gpN5hFy3rf7dTHdEJqlYAul0NoYeUM8B3Epcy1c4nL zTfGoHJIqa+MheVPrSPcOGrztVQSqvd7Lz6l3EoheoeMB9sBEiEggqZXNu2WFdwN0n/Q L1/uRRuHlV6qa90GRTDykd8sCs41XkjP9pJtiOidEfGtc2cqoow19wRpMKXYEvb6Bs6V l0pP+iTRlEecNmYsNvwwIgxV+EffNM5tBF8/OcAFds0m3pGCn2j4ZqRPaih0XrHtJsRN Zb/W72w6ZJA5M3JOiItAdLjBPr87yUgSfAFZJWDsszVDTO282MBbdqMjPtHWT1jIhvKR 5bNQ==
X-Gm-Message-State: AOAM533Ncjg5h3JGnRphqMKeFo4lBy2Xbnz6b0/EvH7yNqNuy0+EgEiv ZrJwaaFTWu5CtIsreHmMIxkel6/YPgt1rFFJbFI=
X-Google-Smtp-Source: ABdhPJxxRnScVuRW39HqSxZr0Ha4XuLG+GAiBqKL10Bc71Uw86+ipDxgBcwr5UI66+jONH4g1khk0FlaIBoMGGxewjA=
X-Received: by 2002:a05:6122:34:: with SMTP id q20mr5032505vkd.66.1591115666654;  Tue, 02 Jun 2020 09:34:26 -0700 (PDT)
MIME-Version: 1.0
References: <159084638843.27466.7915766554130545967@ietfa.amsl.com> <CAKKJt-eHQtgjc-zuO7vrGZ1Q2c7=3hetOb0FyqnEmbTDu1Uwuw@mail.gmail.com>
In-Reply-To: <CAKKJt-eHQtgjc-zuO7vrGZ1Q2c7=3hetOb0FyqnEmbTDu1Uwuw@mail.gmail.com>
From: Matt Joras <matt.joras@gmail.com>
Date: Tue, 2 Jun 2020 09:34:14 -0700
Message-ID: <CADdTf+iBRLu20OH-WTEmo=e7WZ8Ce5QVP+_LWO09u6LxjCPe2g@mail.gmail.com>
Subject: Re: New Version Notification for draft-bonaventure-quic-atsss-overview-00.txt
To: Spencer Dawkins at IETF <spencerdawkins.ietf@gmail.com>
Cc: IETF QUIC WG <quic@ietf.org>
Content-Type: multipart/alternative; boundary="0000000000001b88bb05a71c7cc9"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/FK5w6Cqy5okXzzPbD75GIws3Ukk>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 02 Jun 2020 16:34:57 -0000

--0000000000001b88bb05a71c7cc9
Content-Type: text/plain; charset="UTF-8"

On Tue, Jun 2, 2020 at 4:39 AM Spencer Dawkins at IETF <
spencerdawkins.ietf@gmail.com> wrote:

> Dear QUIC,
>
> Several of us submitted this draft (which I emphasize has no special
> status in IETF, 3GPP, or between IETF and 3GPP).It's being used as one of
> the inputs to SA2#139E, which is underway now (and that also doesn't give
> it any special status)..
>
> We are, of course, interested in comments, especially if we've gotten
> something wrong about QUIC, directions for MP-QUIC, or apparent gaps.
>
> The QUIC chairs have told us this is an appropriate thing to discuss on
> the QUIC mailing list - we did ask.
>
Thank you very much for producing and sharing this document, I agree it is
appropriate for this forum. The diagrams in the document are especially
helpful.

That being said, I have a lot of concerns about the feasibility and
usefulness of the schemes in this document. I wonder if this usage of QUIC
is in the spirit of many QUIC deployments, or if it potentially hinders an
application's usage of QUIC. Correct me if I am wrong, but it seems one of
the core tenets of these schemes is that user applications on mobile
devices would have their flows to servers transparently tunneled over a
multipath transport and then "untunneled" for the final transit to the
server. One of the benefits of QUIC in practice is the ability to
potentially bring the transport implementation "closer" to the application
than was traditionally possible with TCP. One useful signal to applications
is the transport's knowledge of the underlying path state. By transparently
tunneling a QUIC connection over multiple paths this signal is lost. This
has practical implications; the congestion controller used by the transport
cannot proactively act on network switching signals, and the application
cannot proactively adjust its behavior either. As a trivial example, an ABR
streaming video application may detect that the underlying path has
switched and subsequently adjust the bandwidth. One could imagine partially
getting around this problem by implementing a channel on the UE (i.e. some
sort of API) through which the information about the underlying path state
is conveyed to the QUIC transport which is being tunneled. However, this
doesn't address the loss of information for the server. Additionally, I
would argue that if this UE informational channel exists there is no need
to tunnel QUIC traffic at all -- the QUIC transport being utilized directly
by the application can manage its use of the underlying network interfaces.

Said another way, I think it would be useful to further discuss why these
problems need to be solved in "the network" itself rather than focusing on
making QUIC "natively" capable of optimally navigating 5G networks.


>
> Best,
>
> Spencer, but not just Spencer
>
> ---------- Forwarded message ---------
> From: <internet-drafts@ietf.org>
> Date: Sat, May 30, 2020 at 8:46 AM
> Subject: New Version Notification for
> draft-bonaventure-quic-atsss-overview-00.txt
> To: SungHoon Seo <sh.seo@kt.com>, Olivier Bonaventure <
> Olivier.Bonaventure@uclouvain.be>, Qing An <anqing.aq@alibaba-inc.com>,
> Markus Amend <markus.amend@telekom.de>, Andreas Kassler <
> andreas.kassler@kau.se>, Spencer Dawkins <spencerdawkins.ietf@gmail.com>,
> Mohamed Boucadair <mohamed.boucadair@orange.com>, Quentin De Coninck <
> quentin.deconinck@uclouvain.be>, Mirja Kuehlewind <
> mirja.kuehlewind@ericsson.com>, Olivier Bonaventure <
> olivier.bonaventure@uclouvain.be>, Nicolas Keukeleire <
> nicolas.keukeleire@tessares.net>, Maxime Piraux <
> maxime.piraux@uclouvain.be>
>
>
>
> A new version of I-D, draft-bonaventure-quic-atsss-overview-00.txt
> has been successfully submitted by Olivier Bonaventure and posted to the
> IETF repository.
>
> Name:           draft-bonaventure-quic-atsss-overview
> Revision:       00
> Title:          3GPP Access Traffic Steering Switching and Splitting
> (ATSSS) - Overview for IETF Participants
> Document date:  2020-05-30
> Group:          Individual Submission
> Pages:          29
> URL:
> https://www.ietf.org/internet-drafts/draft-bonaventure-quic-atsss-overview-00.txt
> Status:
> https://datatracker.ietf.org/doc/draft-bonaventure-quic-atsss-overview/
> Htmlized:
> https://tools.ietf.org/html/draft-bonaventure-quic-atsss-overview-00
> Htmlized:
> https://datatracker.ietf.org/doc/html/draft-bonaventure-quic-atsss-overview
>
>
> Abstract:
>    This document briefly presents the Access Traffic Steering,
>    Switching, and Splitting (ATSSS) service being specified within the
>    3rd Generation Partnership Project (3GPP).  The ATSSS service
>    provides network support for multihomed devices to select a path for
>    transmission (steer), move traffic from one path to another (switch),
>    or use multiple paths simultaneously (split).  TS 23.501 specifies an
>    ATSSS architecture for TCP traffic.
>
>    This document presents a snap-shot of the ongoing discussion in the
>    3GPP to enable ATSSS for non-TCP traffic, based on the use of QUIC,
>    and assesses to what extent IETF specifications can be used to meet
>    the ATSSS design goals.  Apparent gaps are also documented.
>
>
>
>
> Please note that it may take a couple of minutes from the time of
> submission
> until the htmlized version and diff are available at tools.ietf.org.
>
> The IETF Secretariat
>
>
>
Matt Joras

--0000000000001b88bb05a71c7cc9
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div dir=3D"ltr"><br></div><br><div class=3D"gmail_quote">=
<div dir=3D"ltr" class=3D"gmail_attr">On Tue, Jun 2, 2020 at 4:39 AM Spence=
r Dawkins at IETF &lt;<a href=3D"mailto:spencerdawkins.ietf@gmail.com">spen=
cerdawkins.ietf@gmail.com</a>&gt; wrote:<br></div><blockquote class=3D"gmai=
l_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,20=
4,204);padding-left:1ex"><div dir=3D"ltr">Dear QUIC,=C2=A0<div><br></div><d=
iv>Several of us submitted this draft (which I emphasize has no special sta=
tus in IETF, 3GPP, or between IETF and 3GPP).It&#39;s being used as one of =
the inputs to SA2#139E, which is underway now (and that also doesn&#39;t gi=
ve it any special status)..=C2=A0</div><div><br></div><div>We are, of cours=
e, interested in comments, especially if we&#39;ve gotten something wrong a=
bout QUIC, directions for MP-QUIC, or apparent gaps.</div><div><br></div><d=
iv>The QUIC chairs have told us this is an appropriate thing to discuss on =
the QUIC mailing list - we did ask.</div></div></blockquote><div>Thank you =
very much for producing and sharing this document, I agree it is appropriat=
e for this forum. The diagrams in the document are especially helpful.<br><=
br></div><div>That being said, I have a lot of concerns about the feasibili=
ty and usefulness of the schemes in this document. I wonder if this usage o=
f QUIC is in the spirit of many QUIC deployments, or if it potentially hind=
ers an application&#39;s usage of QUIC. Correct me if I am wrong, but it se=
ems one of the core tenets of these schemes is that user applications on mo=
bile devices would have their flows to servers transparently tunneled over =
a multipath transport and then &quot;untunneled&quot; for the final transit=
 to the server. One of the benefits of QUIC in practice is the ability to p=
otentially bring the transport implementation &quot;closer&quot; to the app=
lication than was traditionally possible with TCP. One useful signal to app=
lications is the transport&#39;s knowledge of the underlying path state. By=
 transparently tunneling a QUIC connection over multiple paths this signal =
is lost. This has practical implications; the congestion controller used by=
 the transport cannot proactively act on network switching signals, and the=
 application cannot proactively adjust its behavior either. As a trivial ex=
ample, an ABR streaming video application may detect that the underlying pa=
th has switched and subsequently adjust the bandwidth. One could imagine pa=
rtially getting around this problem by implementing a channel on the UE (i.=
e. some sort of API) through which the information about the underlying pat=
h state is conveyed to the QUIC transport which is being tunneled. However,=
 this doesn&#39;t address the loss of information for the server. Additiona=
lly, I would argue that if this UE informational channel exists there is no=
 need to tunnel QUIC traffic at all -- the QUIC transport being utilized di=
rectly by the application can manage its use of the underlying network inte=
rfaces.</div><div><br></div><div>Said another way, I think it would be usef=
ul to further discuss why these problems need to be solved in &quot;the net=
work&quot; itself rather than focusing on making QUIC &quot;natively&quot; =
capable of optimally navigating 5G networks.</div><div><br></div><blockquot=
e class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px s=
olid rgb(204,204,204);padding-left:1ex"><div dir=3D"ltr"><div>=C2=A0</div><=
div><br></div><div>Best,</div><div><br></div><div>Spencer, but not just Spe=
ncer<br><br><div class=3D"gmail_quote"><div dir=3D"ltr" class=3D"gmail_attr=
">---------- Forwarded message ---------<br>From: <span dir=3D"auto">&lt;<a=
 href=3D"mailto:internet-drafts@ietf.org" target=3D"_blank">internet-drafts=
@ietf.org</a>&gt;</span><br>Date: Sat, May 30, 2020 at 8:46 AM<br>Subject: =
New Version Notification for draft-bonaventure-quic-atsss-overview-00.txt<b=
r>To: SungHoon Seo &lt;<a href=3D"mailto:sh.seo@kt.com" target=3D"_blank">s=
h.seo@kt.com</a>&gt;, Olivier Bonaventure &lt;<a href=3D"mailto:Olivier.Bon=
aventure@uclouvain.be" target=3D"_blank">Olivier.Bonaventure@uclouvain.be</=
a>&gt;, Qing An &lt;<a href=3D"mailto:anqing.aq@alibaba-inc.com" target=3D"=
_blank">anqing.aq@alibaba-inc.com</a>&gt;, Markus Amend &lt;<a href=3D"mail=
to:markus.amend@telekom.de" target=3D"_blank">markus.amend@telekom.de</a>&g=
t;, Andreas Kassler &lt;<a href=3D"mailto:andreas.kassler@kau.se" target=3D=
"_blank">andreas.kassler@kau.se</a>&gt;, Spencer Dawkins &lt;<a href=3D"mai=
lto:spencerdawkins.ietf@gmail.com" target=3D"_blank">spencerdawkins.ietf@gm=
ail.com</a>&gt;, Mohamed Boucadair &lt;<a href=3D"mailto:mohamed.boucadair@=
orange.com" target=3D"_blank">mohamed.boucadair@orange.com</a>&gt;, Quentin=
 De Coninck &lt;<a href=3D"mailto:quentin.deconinck@uclouvain.be" target=3D=
"_blank">quentin.deconinck@uclouvain.be</a>&gt;, Mirja Kuehlewind &lt;<a hr=
ef=3D"mailto:mirja.kuehlewind@ericsson.com" target=3D"_blank">mirja.kuehlew=
ind@ericsson.com</a>&gt;, Olivier Bonaventure &lt;<a href=3D"mailto:olivier=
.bonaventure@uclouvain.be" target=3D"_blank">olivier.bonaventure@uclouvain.=
be</a>&gt;, Nicolas Keukeleire &lt;<a href=3D"mailto:nicolas.keukeleire@tes=
sares.net" target=3D"_blank">nicolas.keukeleire@tessares.net</a>&gt;, Maxim=
e Piraux &lt;<a href=3D"mailto:maxime.piraux@uclouvain.be" target=3D"_blank=
">maxime.piraux@uclouvain.be</a>&gt;<br></div><br><br><br>
A new version of I-D, draft-bonaventure-quic-atsss-overview-00.txt<br>
has been successfully submitted by Olivier Bonaventure and posted to the<br=
>
IETF repository.<br>
<br>
Name:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0draft-bonaventure-quic-atsss-=
overview<br>
Revision:=C2=A0 =C2=A0 =C2=A0 =C2=A000<br>
Title:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 3GPP Access Traffic Steering Switc=
hing and Splitting (ATSSS) - Overview for IETF Participants<br>
Document date:=C2=A0 2020-05-30<br>
Group:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 Individual Submission<br>
Pages:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 29<br>
URL:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 <a href=3D"https://www.ietf.o=
rg/internet-drafts/draft-bonaventure-quic-atsss-overview-00.txt" rel=3D"nor=
eferrer" target=3D"_blank">https://www.ietf.org/internet-drafts/draft-bonav=
enture-quic-atsss-overview-00.txt</a><br>
Status:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0<a href=3D"https://datatracker.iet=
f.org/doc/draft-bonaventure-quic-atsss-overview/" rel=3D"noreferrer" target=
=3D"_blank">https://datatracker.ietf.org/doc/draft-bonaventure-quic-atsss-o=
verview/</a><br>
Htmlized:=C2=A0 =C2=A0 =C2=A0 =C2=A0<a href=3D"https://tools.ietf.org/html/=
draft-bonaventure-quic-atsss-overview-00" rel=3D"noreferrer" target=3D"_bla=
nk">https://tools.ietf.org/html/draft-bonaventure-quic-atsss-overview-00</a=
><br>
Htmlized:=C2=A0 =C2=A0 =C2=A0 =C2=A0<a href=3D"https://datatracker.ietf.org=
/doc/html/draft-bonaventure-quic-atsss-overview" rel=3D"noreferrer" target=
=3D"_blank">https://datatracker.ietf.org/doc/html/draft-bonaventure-quic-at=
sss-overview</a><br>
<br>
<br>
Abstract:<br>
=C2=A0 =C2=A0This document briefly presents the Access Traffic Steering,<br=
>
=C2=A0 =C2=A0Switching, and Splitting (ATSSS) service being specified withi=
n the<br>
=C2=A0 =C2=A03rd Generation Partnership Project (3GPP).=C2=A0 The ATSSS ser=
vice<br>
=C2=A0 =C2=A0provides network support for multihomed devices to select a pa=
th for<br>
=C2=A0 =C2=A0transmission (steer), move traffic from one path to another (s=
witch),<br>
=C2=A0 =C2=A0or use multiple paths simultaneously (split).=C2=A0 TS 23.501 =
specifies an<br>
=C2=A0 =C2=A0ATSSS architecture for TCP traffic.<br>
<br>
=C2=A0 =C2=A0This document presents a snap-shot of the ongoing discussion i=
n the<br>
=C2=A0 =C2=A03GPP to enable ATSSS for non-TCP traffic, based on the use of =
QUIC,<br>
=C2=A0 =C2=A0and assesses to what extent IETF specifications can be used to=
 meet<br>
=C2=A0 =C2=A0the ATSSS design goals.=C2=A0 Apparent gaps are also documente=
d.<br>
<br>
<br>
<br>
<br>
Please note that it may take a couple of minutes from the time of submissio=
n<br>
until the htmlized version and diff are available at <a href=3D"http://tool=
s.ietf.org" rel=3D"noreferrer" target=3D"_blank">tools.ietf.org</a>.<br>
<br>
The IETF Secretariat<br>
<br>
<br></div></div></div></blockquote><div><br></div><div>Matt Joras <br></div=
></div></div>

--0000000000001b88bb05a71c7cc9--


From nobody Tue Jun  2 12:45:03 2020
Return-Path: <mirja.kuehlewind@ericsson.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 5A5153A0AD7 for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 12:45:01 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.101
X-Spam-Level: 
X-Spam-Status: No, score=-2.101 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, HTML_MESSAGE=0.001, RCVD_IN_MSPIKE_H2=-0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=ericsson.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id DJDTuc_xavQK for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 12:44:57 -0700 (PDT)
Received: from EUR01-HE1-obe.outbound.protection.outlook.com (mail-eopbgr130042.outbound.protection.outlook.com [40.107.13.42]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 751313A0AC7 for <quic@ietf.org>; Tue,  2 Jun 2020 12:44:56 -0700 (PDT)
ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=B10iHx7ughqxvlPavy7BSehsFxNf2AGYSVFzYLuIYtp2jQN5U6JRy5HK7n2CUMXYgCyFRA6PPENXGJlVRTEXwg2QUa2xkBSVP5vdTzEh2y7ASGBp0RKK1thUXFYStfipAS3l5ttRpvKxYRWgr2Huc4pJzUzEpQQSUZlJSH5yO6Nn7stT9e1iM2VkoNkFbsPSvBAqx0L9jLO9+aytZ2mgQsKSTPexYuU411fcuvf1K5RhAC3JHCW5mduzpMffUPIUFwqEU8GVm9nAynO+5ArH+EmvKO8iGCPapD4EiFvmflPPNPBidtI9VXyFNUwkDAKzZ1Y72trdcXEjCVaY12bBIA==
ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com;  s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=6KeDSFBbRDPMZwiMYZMkAVHxECsYcj8FIeK8B0gQJCc=; b=PXXOiMzeG0Hb2mt4ecIE2lCaUQKvrMmBTligtDsAalvTW5Pnn4Cpf22YQZEhq9hFJS8n2ZcS/ycmGw4kTJ2YsNnvewhbu7QWLpBfeKarZnQ3OViih8ar5Fsmq93KNd0eol5RKmQudVM0F4cNwn4ToYRbBCGFrmfLaV0IPk3L2ROBlK+xo2TJDtlQT3TdNyF98ymRXed/KvKqa89MvU4JjfI8rxk6dFLLcEEFlMZ3E0IsGm8nTh84urEp3OUl+JK5kIRDovMh6VNOcTEepi1HI/MULg6EICmqjIb59WuRQf2ml93EVQv0WG6phDq/+T3zz+ggDXxStV38uvsk6MnJsg==
ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=ericsson.com; dmarc=pass action=none header.from=ericsson.com; dkim=pass header.d=ericsson.com; arc=none
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ericsson.com; s=selector1; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=6KeDSFBbRDPMZwiMYZMkAVHxECsYcj8FIeK8B0gQJCc=; b=gWiWXBGG8KRXuGDDcJbFt1TxyCBeJ7kn9J42eLnLgJM5ygHVWllYqUW2VM3qahRr3wNDNw/OODPU7fIPevajVLBi9xnwOourSzjwLz3VrerBczaKq/I8Ub38I96k/B6JAVLgp2wC2c+0oqM4FdqmmzxiJLxaL7kvG0NdxR+qUdo=
Received: from AM0PR07MB4691.eurprd07.prod.outlook.com (2603:10a6:208:75::30) by AM0PR07MB3969.eurprd07.prod.outlook.com (2603:10a6:208:4b::29) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.3066.7; Tue, 2 Jun 2020 19:44:53 +0000
Received: from AM0PR07MB4691.eurprd07.prod.outlook.com ([fe80::5c92:87f1:124a:e912]) by AM0PR07MB4691.eurprd07.prod.outlook.com ([fe80::5c92:87f1:124a:e912%5]) with mapi id 15.20.3066.016; Tue, 2 Jun 2020 19:44:53 +0000
From: Mirja Kuehlewind <mirja.kuehlewind@ericsson.com>
To: Matt Joras <matt.joras@gmail.com>, Spencer Dawkins at IETF <spencerdawkins.ietf@gmail.com>
CC: IETF QUIC WG <quic@ietf.org>
Subject: Re: New Version Notification for draft-bonaventure-quic-atsss-overview-00.txt
Thread-Topic: New Version Notification for draft-bonaventure-quic-atsss-overview-00.txt
Thread-Index: AQHWNoi8vgpf4QkxvU2VNxedroKQQqjFN7mAgABSgwCAAFbLgA==
Date: Tue, 2 Jun 2020 19:44:53 +0000
Message-ID: <D2BBDD3C-89F7-43BF-B5C3-1EC5E8C69EBE@ericsson.com>
References: <159084638843.27466.7915766554130545967@ietfa.amsl.com> <CAKKJt-eHQtgjc-zuO7vrGZ1Q2c7=3hetOb0FyqnEmbTDu1Uwuw@mail.gmail.com> <CADdTf+iBRLu20OH-WTEmo=e7WZ8Ce5QVP+_LWO09u6LxjCPe2g@mail.gmail.com>
In-Reply-To: <CADdTf+iBRLu20OH-WTEmo=e7WZ8Ce5QVP+_LWO09u6LxjCPe2g@mail.gmail.com>
Accept-Language: en-US
Content-Language: en-GB
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
user-agent: Microsoft-MacOutlook/16.37.20051002
authentication-results: gmail.com; dkim=none (message not signed) header.d=none;gmail.com; dmarc=none action=none header.from=ericsson.com;
x-originating-ip: [2003:de:e700:7a00:4b4:966:e26f:f909]
x-ms-publictraffictype: Email
x-ms-office365-filtering-correlation-id: 38e470a4-8642-4361-715a-08d8072d6be7
x-ms-traffictypediagnostic: AM0PR07MB3969:
x-microsoft-antispam-prvs: <AM0PR07MB3969233AF7509882DF82309CF48B0@AM0PR07MB3969.eurprd07.prod.outlook.com>
x-ms-oob-tlc-oobclassifiers: OLM:10000;
x-forefront-prvs: 0422860ED4
x-ms-exchange-senderadcheck: 1
x-microsoft-antispam: BCL:0;
x-microsoft-antispam-message-info: +QVburfOA9RNGsS6j1lM7kR1sU2e3KXxtTfwXt/RwnkaAckxgMBW+mZ3panw1XaNBvmY6oVPknWb6foJesDm/Sv++dvesPBpp/kHxLdznSOII8Tr4hZYDqj7+p9be/92X9gkmmgpqI4rQGnM3Dvo4kWyfKZAm5mImZV58ZB1F85Tlbq3KbGKirxml5TLFhF+tedUvXljk96YsLezFQgyzVzZa5X3Mup9BwULWRal/kh4ndV6JAR98wihUr0aPDcipousnUTajhI5MkBDdi0z7J7L2FSxyoKBVPjUPYznP3xsqhAE47YzbbrSj+5vbjigpu8UvczXfH4Gq+NvP+79+7DHAS/WaFwexPkX7IOV5J+yanVWVfxRg6trJVzYQUCdwiEChfI4zoGQ2KgZXGnVDQ==
x-forefront-antispam-report: CIP:255.255.255.255; CTRY:; LANG:en; SCL:1; SRV:;  IPV:NLI; SFV:NSPM; H:AM0PR07MB4691.eurprd07.prod.outlook.com; PTR:; CAT:NONE;  SFTY:; SFS:(4636009)(396003)(346002)(376002)(366004)(39860400002)(136003)(4326008)(71200400001)(5660300002)(8936002)(8676002)(53546011)(2616005)(110136005)(6506007)(44832011)(316002)(15650500001)(186003)(86362001)(2906002)(83380400001)(33656002)(6486002)(966005)(36756003)(66946007)(66476007)(66556008)(66446008)(166002)(6512007)(76116006)(64756008)(66574014)(478600001); DIR:OUT; SFP:1101; 
x-ms-exchange-antispam-messagedata: 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
x-ms-exchange-transport-forked: True
Content-Type: multipart/alternative; boundary="_000_D2BBDD3C89F743BFB5C31EC5E8C69EBEericssoncom_"
MIME-Version: 1.0
X-OriginatorOrg: ericsson.com
X-MS-Exchange-CrossTenant-Network-Message-Id: 38e470a4-8642-4361-715a-08d8072d6be7
X-MS-Exchange-CrossTenant-originalarrivaltime: 02 Jun 2020 19:44:53.7718 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Hosted
X-MS-Exchange-CrossTenant-id: 92e84ceb-fbfd-47ab-be52-080c6b87953f
X-MS-Exchange-CrossTenant-mailboxtype: HOSTED
X-MS-Exchange-CrossTenant-userprincipalname: jZiy+v9Bf2UtnHFevhAXMq6tiPnCP82+2tDgl0PLmOrYUJcD2wFcaITEnRmeDa1eqSuhK3Dnh//90xBAAZRUq2U4JYRaLUW7gpPCNZYDQDg=
X-MS-Exchange-Transport-CrossTenantHeadersStamped: AM0PR07MB3969
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/JRalB4ezFM5QVL_zmUANqqtd0ao>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 02 Jun 2020 19:45:01 -0000

--_000_D2BBDD3C89F743BFB5C31EC5E8C69EBEericssoncom_
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_D2BBDD3C89F743BFB5C31EC5E8C69EBEericssoncom_
Content-Type: text/html; charset="utf-8"
Content-ID: <1B5745F49713A448B734D1CE8FB03DCB@eurprd07.prod.outlook.com>
Content-Transfer-Encoding: base64
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--_000_D2BBDD3C89F743BFB5C31EC5E8C69EBEericssoncom_--


From nobody Tue Jun  2 13:58:07 2020
Return-Path: <huitema@huitema.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 089BE3A0FE6 for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 13:58:06 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.898
X-Spam-Level: 
X-Spam-Status: No, score=-1.898 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id E4NAzeCRo9d6 for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 13:58:03 -0700 (PDT)
Received: from mx36-out10.antispamcloud.com (mx36-out10.antispamcloud.com [209.126.121.30]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 9BE553A0FE1 for <quic@ietf.org>; Tue,  2 Jun 2020 13:58:03 -0700 (PDT)
Received: from xse381.mail2web.com ([66.113.197.127] helo=xse.mail2web.com) by mx168.antispamcloud.com with esmtp (Exim 4.92) (envelope-from <huitema@huitema.net>) id 1jgDyt-000Oi0-On for quic@ietf.org; Tue, 02 Jun 2020 22:58:02 +0200
Received: from xsmtp21.mail2web.com (unknown [10.100.68.60]) by xse.mail2web.com (Postfix) with ESMTPS id 49c3Mq15dzz1N9H for <quic@ietf.org>; Tue,  2 Jun 2020 13:21:11 -0700 (PDT)
Received: from [10.5.2.18] (helo=xmail08.myhosting.com) by xsmtp21.mail2web.com with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:256) (Exim 4.92) (envelope-from <huitema@huitema.net>) id 1jgDPL-0002Ib-0l for quic@ietf.org; Tue, 02 Jun 2020 13:21:11 -0700
Received: (qmail 23616 invoked from network); 2 Jun 2020 20:21:10 -0000
Received: from unknown (HELO [192.168.1.107]) (Authenticated-user:_huitema@huitema.net@[172.58.43.64]) (envelope-sender <huitema@huitema.net>) by xmail08.myhosting.com (qmail-ldap-1.03) with ESMTPA for <quic@ietf.org>; 2 Jun 2020 20:21:10 -0000
To: Mirja Kuehlewind <mirja.kuehlewind=40ericsson.com@dmarc.ietf.org>, Matt Joras <matt.joras@gmail.com>, Spencer Dawkins at IETF <spencerdawkins.ietf@gmail.com>
Cc: IETF QUIC WG <quic@ietf.org>
References: <159084638843.27466.7915766554130545967@ietfa.amsl.com> <CAKKJt-eHQtgjc-zuO7vrGZ1Q2c7=3hetOb0FyqnEmbTDu1Uwuw@mail.gmail.com> <CADdTf+iBRLu20OH-WTEmo=e7WZ8Ce5QVP+_LWO09u6LxjCPe2g@mail.gmail.com> <D2BBDD3C-89F7-43BF-B5C3-1EC5E8C69EBE@ericsson.com>
From: Christian Huitema <huitema@huitema.net>
Autocrypt: addr=huitema@huitema.net; prefer-encrypt=mutual; keydata= mDMEXtavGxYJKwYBBAHaRw8BAQdA1ou9A5MHTP9N3jfsWzlDZ+jPnQkusmc7sfLmWVz1Rmu0 J0NocmlzdGlhbiBIdWl0ZW1hIDxodWl0ZW1hQGh1aXRlbWEubmV0PoiWBBMWCAA+FiEEw3G4 Nwi4QEpAAXUUELAmqKBYtJQFAl7WrxsCGwMFCQlmAYAFCwkIBwIGFQoJCAsCBBYCAwECHgEC F4AACgkQELAmqKBYtJQbMwD/ebj/qnSbthC/5kD5DxZ/Ip0CGJw5QBz/+fJp3R8iAlsBAMjK r2tmyWyJz0CUkVG24WaR5EAJDvgwDv8h22U6QVkAuDgEXtavGxIKKwYBBAGXVQEFAQEHQJoM 6MUAIqpoqdCIiACiEynZf7nlJg2Eu0pXIhbUGONdAwEIB4h+BBgWCAAmFiEEw3G4Nwi4QEpA AXUUELAmqKBYtJQFAl7WrxsCGwwFCQlmAYAACgkQELAmqKBYtJRm2wD7BzeK5gEXSmBcBf0j BYdSaJcXNzx4yPLbP4GnUMAyl2cBAJzcsR4RkwO4dCRqM9CHpVJCwHtbUDJaa55//E0kp+gH
Subject: Re: New Version Notification for draft-bonaventure-quic-atsss-overview-00.txt
Message-ID: <72be8104-e738-136f-d05c-285fc49533dc@huitema.net>
Date: Tue, 2 Jun 2020 13:21:10 -0700
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:68.0) Gecko/20100101 Thunderbird/68.8.0
MIME-Version: 1.0
In-Reply-To: <D2BBDD3C-89F7-43BF-B5C3-1EC5E8C69EBE@ericsson.com>
Content-Type: multipart/alternative; boundary="------------DB208CD9031F14F4D8EC05F7"
Content-Language: en-US
X-Originating-IP: 66.113.197.127
X-Spampanel-Domain: xsmtpout.mail2web.com
X-Spampanel-Username: 66.113.197.0/24
Authentication-Results: antispamcloud.com; auth=pass smtp.auth=66.113.197.0/24@xsmtpout.mail2web.com
X-Spampanel-Outgoing-Class: unsure
X-Spampanel-Outgoing-Evidence: Combined (0.14)
X-Recommended-Action: accept
X-Filter-ID: Mvzo4OR0dZXEDF/gcnlw0T32zu7lnXf5JHTxKFqDV/OpSDasLI4SayDByyq9LIhV31piCdLkRgHf gOdwi6+6aUTNWdUk1Ol2OGx3IfrIJKywOmJyM1qr8uRnWBrbSAGDcnqpk5VeF3xR4kF6iVwRtbgN zB/4Jkrw1eDLcif59fsTjVhN8wl5j0b4EVC2UUwdU7Tmz6iKnkQL9gqsxD347235Nhqq+/HvroPq 8GSPg+60/QPNqXybIny9WGhadIo/YxFBjiwvt8YGG8KnqhbQXbyme9ldZJ7uNXfg/GfS8fUvP/L5 rCqHDsKZM+xa1iwJX+gRCHfMVnsAk591zk0uilUI+ZL4xWiN8NS6C+dmX6OEdA4u1aThyWrQ/ou2 +v/lmX4Em37yFgrCB6NHRn1g+f3uncIqYSL3lhh5c81YyJqFoLZMmkWsaurVZfvqROaDnDtHb8z5 dpPkEuJ8Snwqla7jUnW3hy14Yji8fo+4xCnSRo4Rcu5Z37rMuDjCny5fE9ykbJ7I9co1MAEE3ruN Xsm8UJsAPvDcVSKtDCYkioPY5Qx4fJOk03R5fJtf/Dv/dkIzS7m4GUpXCY1Y3j3ilZpnfSx1WjM8 Y221ho/pD3/YTgm/NnC/c8YEQ5zoBxO0OPaVMIQAyjx/8zuyzh/o8gJ3dmaBobK9d6GPrTwLmzrv 5pWb82qAoDl3ILGSF0vmDvI0DEihOd7XzCAIcFZdY11677oPXF7r5zsW33ZNliqbq1iQQhopQxqA dW22RV240rHJtSVPlENYbD9JAuv2lDzAXPU1UvCyITFs2AML0Poh8JzgkVH7m09sItCDRNTRHHAS JNUmoOHSoqgqxfHmWRWcrRhLeB34s3hUb32GO+06ejVsMqjCBuJMyJeyF7Ye08QV3No+S2msRDep v5w/kkG0v17AmegcpQ0tml/sN9lmMy/o83jVXTcfb9k0nLWblJy7uxV6dw8jzlsaNZe6hynMJcjx DydxsJEju76A7X1QIVydqXpZ6MHhiKws9Iiut28r9wo4SqUIg8Yh9hAM0n3LLzx/F2gT3wl8JQJv Bho=
X-Report-Abuse-To: spam@quarantine11.antispamcloud.com
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/5rkW9BQTGAfrK1gaoEyTjYXTD3o>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 02 Jun 2020 20:58:06 -0000

This is a multi-part message in MIME format.
--------------DB208CD9031F14F4D8EC05F7
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable

I think that we should be very clear that the primary goal for QUIC
multipath design is end-to-end. For example, a device might very well
have connection with 2 cellular networks, each providing their own
"advanced" services, and use end-to-end multipath to balance traffic
between the two networks.

We also have to consider multipath privacy. In the ATSSS setup, the
device is supposed to establish a connection with the mobile operator's
service even in cases when it is only connected to a Wi-Fi network. That
means disclosing connectivity and activity to the operator, just in case
a backup through cellular might be needed. This is definitely a
trade-off that many applications will not want to make. End to end
multipath alleviates that concern, because the network operators only
learn of the device's activity when the devices actually uses their netwo=
rk.

-- Christian Huitema

On 6/2/2020 12:44 PM, Mirja Kuehlewind wrote:
>
> Hi Matt,
>
> =C2=A0
>
> thanks for the feedback. Please note that in ATSSS the UE can decide
> to use this multipath service, or not, e.g. if the server provides
> multipath support directly (or even do both but not sure how useful
> that is). However, using the network supported service can be
> beneficial for downlink traffic towards the UE as the network might
> have information about network state and potential needs to handover.
> This information can be shared with the UE, however, there is no
> protocol mechanism to share this information with the server. I
> believe we briefly mention this in the draft but I guess this
> discussion could definitely be more explicit.
>
> =C2=A0
>
> Mirja
>
> =C2=A0
>
> =C2=A0
>
> =C2=A0
>
> *From: *QUIC <quic-bounces@ietf.org> on behalf of Matt Joras
> <matt.joras@gmail.com>
> *Date: *Tuesday, 2. June 2020 at 18:36
> *To: *Spencer Dawkins at IETF <spencerdawkins.ietf@gmail.com>
> *Cc: *IETF QUIC WG <quic@ietf.org>
> *Subject: *Re: New Version Notification for
> draft-bonaventure-quic-atsss-overview-00.txt
>
> =C2=A0
>
> =C2=A0
>
> =C2=A0
>
> On Tue, Jun 2, 2020 at 4:39 AM Spencer Dawkins at IETF
> <spencerdawkins.ietf@gmail.com <mailto:spencerdawkins.ietf@gmail.com>>
> wrote:
>
>     Dear QUIC,=C2=A0
>
>     =C2=A0
>
>     Several of us submitted this draft (which I emphasize has no
>     special status in IETF, 3GPP, or between IETF and 3GPP).It's being
>     used as one of the inputs to SA2#139E, which is underway now (and
>     that also doesn't give it any special status)..=C2=A0
>
>     =C2=A0
>
>     We are, of course, interested in comments, especially if we've
>     gotten something wrong about QUIC, directions for MP-QUIC, or
>     apparent gaps.
>
>     =C2=A0
>
>     The QUIC chairs have told us this is an appropriate thing to
>     discuss on the QUIC mailing list - we did ask.
>
> Thank you very much for producing and sharing this document, I agree
> it is appropriate for this forum. The diagrams in the document are
> especially helpful.
>
> That being said, I have a lot of concerns about the feasibility and
> usefulness of the schemes in this document. I wonder if this usage of
> QUIC is in the spirit of many QUIC deployments, or if it potentially
> hinders an application's usage of QUIC. Correct me if I am wrong, but
> it seems one of the core tenets of these schemes is that user
> applications on mobile devices would have their flows to servers
> transparently tunneled over a multipath transport and then
> "untunneled" for the final transit to the server. One of the benefits
> of QUIC in practice is the ability to potentially bring the transport
> implementation "closer" to the application than was traditionally
> possible with TCP. One useful signal to applications is the
> transport's knowledge of the underlying path state. By transparently
> tunneling a QUIC connection over multiple paths this signal is lost.
> This has practical implications; the congestion controller used by the
> transport cannot proactively act on network switching signals, and the
> application cannot proactively adjust its behavior either. As a
> trivial example, an ABR streaming video application may detect that
> the underlying path has switched and subsequently adjust the
> bandwidth. One could imagine partially getting around this problem by
> implementing a channel on the UE (i.e. some sort of API) through which
> the information about the underlying path state is conveyed to the
> QUIC transport which is being tunneled. However, this doesn't address
> the loss of information for the server. Additionally, I would argue
> that if this UE informational channel exists there is no need to
> tunnel QUIC traffic at all -- the QUIC transport being utilized
> directly by the application can manage its use of the underlying
> network interfaces.
>
> =C2=A0
>
> Said another way, I think it would be useful to further discuss why
> these problems need to be solved in "the network" itself rather than
> focusing on making QUIC "natively" capable of optimally navigating 5G
> networks.
>
> =C2=A0
>
>     =C2=A0
>
>     =C2=A0
>
>     Best,
>
>     =C2=A0
>
>     Spencer, but not just Spencer
>
>     ---------- Forwarded message ---------
>     From: <internet-drafts@ietf.org <mailto:internet-drafts@ietf.org>>
>     Date: Sat, May 30, 2020 at 8:46 AM
>     Subject: New Version Notification for
>     draft-bonaventure-quic-atsss-overview-00.txt
>     To: SungHoon Seo <sh.seo@kt.com <mailto:sh.seo@kt.com>>, Olivier
>     Bonaventure <Olivier.Bonaventure@uclouvain.be
>     <mailto:Olivier.Bonaventure@uclouvain.be>>, Qing An
>     <anqing.aq@alibaba-inc.com <mailto:anqing.aq@alibaba-inc.com>>,
>     Markus Amend <markus.amend@telekom.de
>     <mailto:markus.amend@telekom.de>>, Andreas Kassler
>     <andreas.kassler@kau.se <mailto:andreas.kassler@kau.se>>, Spencer
>     Dawkins <spencerdawkins.ietf@gmail.com
>     <mailto:spencerdawkins.ietf@gmail.com>>, Mohamed Boucadair
>     <mohamed.boucadair@orange.com
>     <mailto:mohamed.boucadair@orange.com>>, Quentin De Coninck
>     <quentin.deconinck@uclouvain.be
>     <mailto:quentin.deconinck@uclouvain.be>>, Mirja Kuehlewind
>     <mirja.kuehlewind@ericsson.com
>     <mailto:mirja.kuehlewind@ericsson.com>>, Olivier Bonaventure
>     <olivier.bonaventure@uclouvain.be
>     <mailto:olivier..bonaventure@uclouvain.be>>, Nicolas Keukeleire
>     <nicolas.keukeleire@tessares.net
>     <mailto:nicolas.keukeleire@tessares.net>>, Maxime Piraux
>     <maxime.piraux@uclouvain.be <mailto:maxime.piraux@uclouvain.be>>
>
>
>
>
>     A new version of I-D, draft-bonaventure-quic-atsss-overview-00.txt
>     has been successfully submitted by Olivier Bonaventure and posted
>     to the
>     IETF repository.
>
>     Name:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0draft-bonaventure-qui=
c-atsss-overview
>     Revision:=C2=A0 =C2=A0 =C2=A0 =C2=A000
>     Title:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 3GPP Access Traffic Steeri=
ng Switching and
>     Splitting (ATSSS) - Overview for IETF Participants
>     Document date:=C2=A0 2020-05-30
>     Group:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 Individual Submission
>     Pages:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 29
>     URL:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0
>     https://www.ietf.org/internet-drafts/draft-bonaventure-quic-atsss-o=
verview-00.txt
>     Status:=C2=A0 =C2=A0 =C2=A0 =C2=A0
>     =C2=A0https://datatracker.ietf.org/doc/draft-bonaventure-quic-atsss=
-overview/
>     Htmlized:=C2=A0 =C2=A0 =C2=A0
>     =C2=A0https://tools.ietf.org/html/draft-bonaventure-quic-atsss-over=
view-00
>     Htmlized:=C2=A0 =C2=A0 =C2=A0
>     =C2=A0https://datatracker.ietf.org/doc/html/draft-bonaventure-quic-=
atsss-overview
>
>
>     Abstract:
>     =C2=A0 =C2=A0This document briefly presents the Access Traffic Stee=
ring,
>     =C2=A0 =C2=A0Switching, and Splitting (ATSSS) service being specifi=
ed within the
>     =C2=A0 =C2=A03rd Generation Partnership Project (3GPP).=C2=A0 The A=
TSSS service
>     =C2=A0 =C2=A0provides network support for multihomed devices to sel=
ect a
>     path for
>     =C2=A0 =C2=A0transmission (steer), move traffic from one path to an=
other
>     (switch),
>     =C2=A0 =C2=A0or use multiple paths simultaneously (split).=C2=A0 TS=
 23.501
>     specifies an
>     =C2=A0 =C2=A0ATSSS architecture for TCP traffic.
>
>     =C2=A0 =C2=A0This document presents a snap-shot of the ongoing disc=
ussion in the
>     =C2=A0 =C2=A03GPP to enable ATSSS for non-TCP traffic, based on the=
 use of QUIC,
>     =C2=A0 =C2=A0and assesses to what extent IETF specifications can be=
 used to meet
>     =C2=A0 =C2=A0the ATSSS design goals.=C2=A0 Apparent gaps are also d=
ocumented.
>
>
>
>
>     Please note that it may take a couple of minutes from the time of
>     submission
>     until the htmlized version and diff are available at
>     tools.ietf.org <http://tools.ietf.org>.
>
>     The IETF Secretariat
>
> =C2=A0
>
> Matt Joras
>

--------------DB208CD9031F14F4D8EC05F7
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: 8bit

<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <p>I think that we should be very clear that the primary goal for
      QUIC multipath design is end-to-end. For example, a device might
      very well have connection with 2 cellular networks, each providing
      their own "advanced" services, and use end-to-end multipath to
      balance traffic between the two networks.</p>
    <p>We also have to consider multipath privacy. In the ATSSS setup,
      the device is supposed to establish a connection with the mobile
      operator's service even in cases when it is only connected to a
      Wi-Fi network. That means disclosing connectivity and activity to
      the operator, just in case a backup through cellular might be
      needed. This is definitely a trade-off that many applications will
      not want to make. End to end multipath alleviates that concern,
      because the network operators only learn of the device's activity
      when the devices actually uses their network.</p>
    <p>-- Christian Huitema<br>
    </p>
    <div class="moz-cite-prefix">On 6/2/2020 12:44 PM, Mirja Kuehlewind
      wrote:<br>
    </div>
    <blockquote type="cite"
      cite="mid:D2BBDD3C-89F7-43BF-B5C3-1EC5E8C69EBE@ericsson.com">
      <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
      <meta name="Generator" content="Microsoft Word 15 (filtered
        medium)">
      <style><!--
/* Font Definitions */
@font-face
	{font-family:"Cambria Math";
	panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
	{font-family:Calibri;
	panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
	{margin:0cm;
	margin-bottom:.0001pt;
	font-size:11.0pt;
	font-family:"Calibri",sans-serif;}
a:link, span.MsoHyperlink
	{mso-style-priority:99;
	color:blue;
	text-decoration:underline;}
span.EmailStyle18
	{mso-style-type:personal-reply;
	font-family:"Calibri",sans-serif;
	color:windowtext;}
.MsoChpDefault
	{mso-style-type:export-only;
	font-size:10.0pt;}
@page WordSection1
	{size:612.0pt 792.0pt;
	margin:72.0pt 72.0pt 72.0pt 72.0pt;}
div.WordSection1
	{page:WordSection1;}
--></style>
      <div class="WordSection1">
        <p class="MsoNormal"><span style="mso-fareast-language:EN-US"
            lang="EN-US">Hi Matt,<o:p></o:p></span></p>
        <p class="MsoNormal"><span style="mso-fareast-language:EN-US"
            lang="EN-US"><o:p> </o:p></span></p>
        <p class="MsoNormal"><span style="mso-fareast-language:EN-US"
            lang="EN-US">thanks for the feedback. Please note that in
            ATSSS the UE can decide to use this multipath service, or
            not, e.g. if the server provides multipath support directly
            (or even do both but not sure how useful that is). However,
            using the network supported service can be beneficial for
            downlink traffic towards the UE as the network might have
            information about network state and potential needs to
            handover. This information can be shared with the UE,
            however, there is no protocol mechanism to share this
            information with the server. I believe we briefly mention
            this in the draft but I guess this discussion could
            definitely be more explicit.
            <o:p></o:p></span></p>
        <p class="MsoNormal"><span style="mso-fareast-language:EN-US"
            lang="EN-US"><o:p> </o:p></span></p>
        <p class="MsoNormal"><span style="mso-fareast-language:EN-US"
            lang="EN-US">Mirja<o:p></o:p></span></p>
        <p class="MsoNormal"><span style="mso-fareast-language:EN-US"
            lang="EN-US"><o:p> </o:p></span></p>
        <p class="MsoNormal"><span style="mso-fareast-language:EN-US"
            lang="EN-US"><o:p> </o:p></span></p>
        <p class="MsoNormal"><span style="mso-fareast-language:EN-US"><o:p> </o:p></span></p>
        <div style="border:none;border-top:solid #B5C4DF
          1.0pt;padding:3.0pt 0cm 0cm 0cm">
          <p class="MsoNormal" style="margin-left:36.0pt"><b><span
                style="font-size:12.0pt;color:black">From:
              </span></b><span style="font-size:12.0pt;color:black">QUIC
              <a class="moz-txt-link-rfc2396E" href="mailto:quic-bounces@ietf.org">&lt;quic-bounces@ietf.org&gt;</a> on behalf of Matt Joras
              <a class="moz-txt-link-rfc2396E" href="mailto:matt.joras@gmail.com">&lt;matt.joras@gmail.com&gt;</a><br>
              <b>Date: </b>Tuesday, 2. June 2020 at 18:36<br>
              <b>To: </b>Spencer Dawkins at IETF
              <a class="moz-txt-link-rfc2396E" href="mailto:spencerdawkins.ietf@gmail.com">&lt;spencerdawkins.ietf@gmail.com&gt;</a><br>
              <b>Cc: </b>IETF QUIC WG <a class="moz-txt-link-rfc2396E" href="mailto:quic@ietf.org">&lt;quic@ietf.org&gt;</a><br>
              <b>Subject: </b>Re: New Version Notification for
              draft-bonaventure-quic-atsss-overview-00.txt<o:p></o:p></span></p>
        </div>
        <div>
          <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
        </div>
        <div>
          <div>
            <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
          </div>
          <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
          <div>
            <div>
              <p class="MsoNormal" style="margin-left:36.0pt">On Tue,
                Jun 2, 2020 at 4:39 AM Spencer Dawkins at IETF &lt;<a
                  href="mailto:spencerdawkins.ietf@gmail.com"
                  moz-do-not-send="true">spencerdawkins.ietf@gmail.com</a>&gt;
                wrote:<o:p></o:p></p>
            </div>
            <blockquote style="border:none;border-left:solid #CCCCCC
              1.0pt;padding:0cm 0cm 0cm
              6.0pt;margin-left:4.8pt;margin-right:0cm">
              <div>
                <p class="MsoNormal" style="margin-left:36.0pt">Dear
                  QUIC, <o:p></o:p></p>
                <div>
                  <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
                </div>
                <div>
                  <p class="MsoNormal" style="margin-left:36.0pt">Several
                    of us submitted this draft (which I emphasize has no
                    special status in IETF, 3GPP, or between IETF and
                    3GPP).It's being used as one of the inputs to
                    SA2#139E, which is underway now (and that also
                    doesn't give it any special status).. <o:p></o:p></p>
                </div>
                <div>
                  <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
                </div>
                <div>
                  <p class="MsoNormal" style="margin-left:36.0pt">We
                    are, of course, interested in comments, especially
                    if we've gotten something wrong about QUIC,
                    directions for MP-QUIC, or apparent gaps.<o:p></o:p></p>
                </div>
                <div>
                  <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
                </div>
                <div>
                  <p class="MsoNormal" style="margin-left:36.0pt">The
                    QUIC chairs have told us this is an appropriate
                    thing to discuss on the QUIC mailing list - we did
                    ask.<o:p></o:p></p>
                </div>
              </div>
            </blockquote>
            <div>
              <p class="MsoNormal"
style="mso-margin-top-alt:0cm;margin-right:0cm;margin-bottom:12.0pt;margin-left:36.0pt">Thank
                you very much for producing and sharing this document, I
                agree it is appropriate for this forum. The diagrams in
                the document are especially helpful.<o:p></o:p></p>
            </div>
            <div>
              <p class="MsoNormal" style="margin-left:36.0pt">That being
                said, I have a lot of concerns about the feasibility and
                usefulness of the schemes in this document. I wonder if
                this usage of QUIC is in the spirit of many QUIC
                deployments, or if it potentially hinders an
                application's usage of QUIC. Correct me if I am wrong,
                but it seems one of the core tenets of these schemes is
                that user applications on mobile devices would have
                their flows to servers transparently tunneled over a
                multipath transport and then "untunneled" for the final
                transit to the server. One of the benefits of QUIC in
                practice is the ability to potentially bring the
                transport implementation "closer" to the application
                than was traditionally possible with TCP. One useful
                signal to applications is the transport's knowledge of
                the underlying path state. By transparently tunneling a
                QUIC connection over multiple paths this signal is lost.
                This has practical implications; the congestion
                controller used by the transport cannot proactively act
                on network switching signals, and the application cannot
                proactively adjust its behavior either. As a trivial
                example, an ABR streaming video application may detect
                that the underlying path has switched and subsequently
                adjust the bandwidth. One could imagine partially
                getting around this problem by implementing a channel on
                the UE (i.e. some sort of API) through which the
                information about the underlying path state is conveyed
                to the QUIC transport which is being tunneled. However,
                this doesn't address the loss of information for the
                server. Additionally, I would argue that if this UE
                informational channel exists there is no need to tunnel
                QUIC traffic at all -- the QUIC transport being utilized
                directly by the application can manage its use of the
                underlying network interfaces.<o:p></o:p></p>
            </div>
            <div>
              <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
            </div>
            <div>
              <p class="MsoNormal" style="margin-left:36.0pt">Said
                another way, I think it would be useful to further
                discuss why these problems need to be solved in "the
                network" itself rather than focusing on making QUIC
                "natively" capable of optimally navigating 5G networks.<o:p></o:p></p>
            </div>
            <div>
              <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
            </div>
            <blockquote style="border:none;border-left:solid #CCCCCC
              1.0pt;padding:0cm 0cm 0cm
              6.0pt;margin-left:4.8pt;margin-right:0cm">
              <div>
                <div>
                  <p class="MsoNormal" style="margin-left:36.0pt"> <o:p></o:p></p>
                </div>
                <div>
                  <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
                </div>
                <div>
                  <p class="MsoNormal" style="margin-left:36.0pt">Best,<o:p></o:p></p>
                </div>
                <div>
                  <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
                </div>
                <div>
                  <p class="MsoNormal"
style="mso-margin-top-alt:0cm;margin-right:0cm;margin-bottom:12.0pt;margin-left:36.0pt">Spencer,
                    but not just Spencer<o:p></o:p></p>
                  <div>
                    <div>
                      <p class="MsoNormal" style="margin-left:36.0pt">----------
                        Forwarded message ---------<br>
                        From: &lt;<a
                          href="mailto:internet-drafts@ietf.org"
                          target="_blank" moz-do-not-send="true">internet-drafts@ietf.org</a>&gt;<br>
                        Date: Sat, May 30, 2020 at 8:46 AM<br>
                        Subject: New Version Notification for
                        draft-bonaventure-quic-atsss-overview-00.txt<br>
                        To: SungHoon Seo &lt;<a
                          href="mailto:sh.seo@kt.com" target="_blank"
                          moz-do-not-send="true">sh.seo@kt.com</a>&gt;,
                        Olivier Bonaventure &lt;<a
                          href="mailto:Olivier.Bonaventure@uclouvain.be"
                          target="_blank" moz-do-not-send="true">Olivier.Bonaventure@uclouvain.be</a>&gt;,
                        Qing An &lt;<a
                          href="mailto:anqing.aq@alibaba-inc.com"
                          target="_blank" moz-do-not-send="true">anqing.aq@alibaba-inc.com</a>&gt;,
                        Markus Amend &lt;<a
                          href="mailto:markus.amend@telekom.de"
                          target="_blank" moz-do-not-send="true">markus.amend@telekom.de</a>&gt;,
                        Andreas Kassler &lt;<a
                          href="mailto:andreas.kassler@kau.se"
                          target="_blank" moz-do-not-send="true">andreas.kassler@kau.se</a>&gt;,
                        Spencer Dawkins &lt;<a
                          href="mailto:spencerdawkins.ietf@gmail.com"
                          target="_blank" moz-do-not-send="true">spencerdawkins.ietf@gmail.com</a>&gt;,
                        Mohamed Boucadair &lt;<a
                          href="mailto:mohamed.boucadair@orange.com"
                          target="_blank" moz-do-not-send="true">mohamed.boucadair@orange.com</a>&gt;,
                        Quentin De Coninck &lt;<a
                          href="mailto:quentin.deconinck@uclouvain.be"
                          target="_blank" moz-do-not-send="true">quentin.deconinck@uclouvain.be</a>&gt;,
                        Mirja Kuehlewind &lt;<a
                          href="mailto:mirja.kuehlewind@ericsson.com"
                          target="_blank" moz-do-not-send="true">mirja.kuehlewind@ericsson.com</a>&gt;,
                        Olivier Bonaventure &lt;<a
                          href="mailto:olivier..bonaventure@uclouvain.be"
                          target="_blank" moz-do-not-send="true">olivier.bonaventure@uclouvain.be</a>&gt;,
                        Nicolas Keukeleire &lt;<a
                          href="mailto:nicolas.keukeleire@tessares.net"
                          target="_blank" moz-do-not-send="true">nicolas.keukeleire@tessares.net</a>&gt;,
                        Maxime Piraux &lt;<a
                          href="mailto:maxime.piraux@uclouvain.be"
                          target="_blank" moz-do-not-send="true">maxime.piraux@uclouvain.be</a>&gt;<o:p></o:p></p>
                    </div>
                    <p class="MsoNormal"
style="mso-margin-top-alt:0cm;margin-right:0cm;margin-bottom:12.0pt;margin-left:36.0pt"><br>
                      <br>
                      <br>
                      A new version of I-D,
                      draft-bonaventure-quic-atsss-overview-00.txt<br>
                      has been successfully submitted by Olivier
                      Bonaventure and posted to the<br>
                      IETF repository.<br>
                      <br>
                      Name:         
                       draft-bonaventure-quic-atsss-overview<br>
                      Revision:       00<br>
                      Title:          3GPP Access Traffic Steering
                      Switching and Splitting (ATSSS) - Overview for
                      IETF Participants<br>
                      Document date:  2020-05-30<br>
                      Group:          Individual Submission<br>
                      Pages:          29<br>
                      URL:            <a
href="https://www.ietf.org/internet-drafts/draft-bonaventure-quic-atsss-overview-00.txt"
                        target="_blank" moz-do-not-send="true">
https://www.ietf.org/internet-drafts/draft-bonaventure-quic-atsss-overview-00.txt</a><br>
                      Status:         <a
href="https://datatracker.ietf.org/doc/draft-bonaventure-quic-atsss-overview/"
                        target="_blank" moz-do-not-send="true">https://datatracker.ietf.org/doc/draft-bonaventure-quic-atsss-overview/</a><br>
                      Htmlized:       <a
href="https://tools.ietf.org/html/draft-bonaventure-quic-atsss-overview-00"
                        target="_blank" moz-do-not-send="true">https://tools.ietf.org/html/draft-bonaventure-quic-atsss-overview-00</a><br>
                      Htmlized:       <a
href="https://datatracker.ietf.org/doc/html/draft-bonaventure-quic-atsss-overview"
                        target="_blank" moz-do-not-send="true">https://datatracker.ietf.org/doc/html/draft-bonaventure-quic-atsss-overview</a><br>
                      <br>
                      <br>
                      Abstract:<br>
                         This document briefly presents the Access
                      Traffic Steering,<br>
                         Switching, and Splitting (ATSSS) service being
                      specified within the<br>
                         3rd Generation Partnership Project (3GPP).  The
                      ATSSS service<br>
                         provides network support for multihomed devices
                      to select a path for<br>
                         transmission (steer), move traffic from one
                      path to another (switch),<br>
                         or use multiple paths simultaneously (split). 
                      TS 23.501 specifies an<br>
                         ATSSS architecture for TCP traffic.<br>
                      <br>
                         This document presents a snap-shot of the
                      ongoing discussion in the<br>
                         3GPP to enable ATSSS for non-TCP traffic, based
                      on the use of QUIC,<br>
                         and assesses to what extent IETF specifications
                      can be used to meet<br>
                         the ATSSS design goals.  Apparent gaps are also
                      documented.<br>
                      <br>
                      <br>
                      <br>
                      <br>
                      Please note that it may take a couple of minutes
                      from the time of submission<br>
                      until the htmlized version and diff are available
                      at <a href="http://tools.ietf.org"
                        target="_blank" moz-do-not-send="true">
                        tools.ietf.org</a>.<br>
                      <br>
                      The IETF Secretariat<br>
                      <br>
                      <o:p></o:p></p>
                  </div>
                </div>
              </div>
            </blockquote>
            <div>
              <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
            </div>
            <div>
              <p class="MsoNormal" style="margin-left:36.0pt">Matt Joras
                <o:p></o:p></p>
            </div>
          </div>
        </div>
      </div>
    </blockquote>
  </body>
</html>

--------------DB208CD9031F14F4D8EC05F7--


From nobody Tue Jun  2 15:19:25 2020
Return-Path: <roland@zinks.de>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 5C5D23A107A for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 15:19:24 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.098
X-Spam-Level: 
X-Spam-Status: No, score=-2.098 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, HTML_MESSAGE=0.001, RCVD_IN_MSPIKE_H2=-0.001, SPF_NONE=0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=zinks.de
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id bvq0KKoEdVeq for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 15:19:20 -0700 (PDT)
Received: from mo4-p00-ob.smtp.rzone.de (mo4-p00-ob.smtp.rzone.de [81.169.146.160]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 78E2B3A1079 for <quic@ietf.org>; Tue,  2 Jun 2020 15:19:19 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; t=1591136349; s=strato-dkim-0002; d=zinks.de; h=In-Reply-To:Date:Message-ID:From:References:To:Subject: X-RZG-CLASS-ID:X-RZG-AUTH:From:Subject:Sender; bh=4Ei1WcjjoKv49Ksq9coSzcC0bw2wrfgKCWDI/3gZ4gg=; b=ZZItwiPJYNCw2d2a03TbnZO7/sJvE5Ns08TYu2JZ7TkuDqgcU1Yo2u450fOmmp4GcQ X9gaqyhhpG76qOH1Wve5EAQmqPGAnGnFJ6HNsaewgzo/jeH8JflpPWgOAXBiTPr0Lccz S81/G1X324dN3UOqmvQKfFRXuppsPfIy+ETyiIEEaLT99uxhedVTew5EJjGXcDMyHEFe +kB/dAjjh4weRnJ1/kezCw8ZUwzVIhI5ID5B5GCwWwTJRqXqabRYryDHoH2nPRtx3sbI GrzmgGEf0tiYvRpDwao+SyOgmhuwt/Xtt3cszNQ/kAX24BwW4z9B8e7fj2kh5a2TjuaI 6LfQ==
X-RZG-AUTH: ":PmMIdE6sW+WWP9q/oR3Lt+I+9LAZzXrcq8knhvfmBiJzkmKn1YaZ2OgvlDcDHae29xg="
X-RZG-CLASS-ID: mo00
Received: from [10.10.10.105] by smtp.strato.de (RZmta 46.9.0 DYNA|AUTH) with ESMTPSA id L02360w52MJ9Ml6 (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256 bits)) (Client did not present a certificate) for <quic@ietf.org>; Wed, 3 Jun 2020 00:19:09 +0200 (CEST)
Subject: Re: New Version Notification for draft-bonaventure-quic-atsss-overview-00.txt
To: quic@ietf.org
References: <159084638843.27466.7915766554130545967@ietfa.amsl.com> <CAKKJt-eHQtgjc-zuO7vrGZ1Q2c7=3hetOb0FyqnEmbTDu1Uwuw@mail.gmail.com> <CADdTf+iBRLu20OH-WTEmo=e7WZ8Ce5QVP+_LWO09u6LxjCPe2g@mail.gmail.com> <D2BBDD3C-89F7-43BF-B5C3-1EC5E8C69EBE@ericsson.com> <72be8104-e738-136f-d05c-285fc49533dc@huitema.net>
From: Roland Zink <roland@zinks.de>
Message-ID: <f2d20d2f-d574-e0bb-cd71-35503964c82c@zinks.de>
Date: Wed, 3 Jun 2020 00:19:08 +0200
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:68.0) Gecko/20100101 Thunderbird/68.8.1
MIME-Version: 1.0
In-Reply-To: <72be8104-e738-136f-d05c-285fc49533dc@huitema.net>
Content-Type: multipart/alternative; boundary="------------C3B0C642C2728766012F67C5"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/drDXwduCm_ZqSLUET8qgTRxZj4s>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 02 Jun 2020 22:19:24 -0000

This is a multi-part message in MIME format.
--------------C3B0C642C2728766012F67C5
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 8bit

Hi Christian,


When I understood this right for MPTCP then this can be applied 
end-to-end without any ATSSS interaction. The ATSSS proxy is only needed 
when the end servers do not support MPTCP. When there is end-to-end 
MPQUIC then there doesn't seem to be a need for ATSSS as well.


A cellular network often knows something about mobile devices even when 
those devices are not actively using it. A device registers with a 
cellular network and gets an IP address. After that the cellular network 
keeps track of the device location area (only an area, not the exact 
location) even when the device is inactive or connected to WiFi. This is 
done in case some data destinated to the device (IP) arrives in the 
cellular network (PGW/UPF). Then the device is paged in the area and the 
data is delivered after the device becomes active. You can of cause turn 
off cellular if you don't want this.


Regards,

Roland


Am 02.06.2020 um 22:21 schrieb Christian Huitema:
>
> I think that we should be very clear that the primary goal for QUIC 
> multipath design is end-to-end. For example, a device might very well 
> have connection with 2 cellular networks, each providing their own 
> "advanced" services, and use end-to-end multipath to balance traffic 
> between the two networks.
>
> We also have to consider multipath privacy. In the ATSSS setup, the 
> device is supposed to establish a connection with the mobile 
> operator's service even in cases when it is only connected to a Wi-Fi 
> network. That means disclosing connectivity and activity to the 
> operator, just in case a backup through cellular might be needed. This 
> is definitely a trade-off that many applications will not want to 
> make. End to end multipath alleviates that concern, because the 
> network operators only learn of the device's activity when the devices 
> actually uses their network.
>
> -- Christian Huitema
>
> On 6/2/2020 12:44 PM, Mirja Kuehlewind wrote:
>>
>> Hi Matt,
>>
>> thanks for the feedback. Please note that in ATSSS the UE can decide 
>> to use this multipath service, or not, e.g. if the server provides 
>> multipath support directly (or even do both but not sure how useful 
>> that is). However, using the network supported service can be 
>> beneficial for downlink traffic towards the UE as the network might 
>> have information about network state and potential needs to handover. 
>> This information can be shared with the UE, however, there is no 
>> protocol mechanism to share this information with the server. I 
>> believe we briefly mention this in the draft but I guess this 
>> discussion could definitely be more explicit.
>>
>> Mirja
>>
>> *From: *QUIC <quic-bounces@ietf.org> on behalf of Matt Joras 
>> <matt.joras@gmail.com>
>> *Date: *Tuesday, 2. June 2020 at 18:36
>> *To: *Spencer Dawkins at IETF <spencerdawkins.ietf@gmail.com>
>> *Cc: *IETF QUIC WG <quic@ietf.org>
>> *Subject: *Re: New Version Notification for 
>> draft-bonaventure-quic-atsss-overview-00.txt
>>
>> On Tue, Jun 2, 2020 at 4:39 AM Spencer Dawkins at IETF 
>> <spencerdawkins.ietf@gmail.com 
>> <mailto:spencerdawkins.ietf@gmail.com>> wrote:
>>
>>     Dear QUIC,
>>
>>     Several of us submitted this draft (which I emphasize has no
>>     special status in IETF, 3GPP, or between IETF and 3GPP).It's
>>     being used as one of the inputs to SA2#139E, which is underway
>>     now (and that also doesn't give it any special status)..
>>
>>     We are, of course, interested in comments, especially if we've
>>     gotten something wrong about QUIC, directions for MP-QUIC, or
>>     apparent gaps.
>>
>>     The QUIC chairs have told us this is an appropriate thing to
>>     discuss on the QUIC mailing list - we did ask.
>>
>> Thank you very much for producing and sharing this document, I agree 
>> it is appropriate for this forum. The diagrams in the document are 
>> especially helpful.
>>
>> That being said, I have a lot of concerns about the feasibility and 
>> usefulness of the schemes in this document. I wonder if this usage of 
>> QUIC is in the spirit of many QUIC deployments, or if it potentially 
>> hinders an application's usage of QUIC. Correct me if I am wrong, but 
>> it seems one of the core tenets of these schemes is that user 
>> applications on mobile devices would have their flows to servers 
>> transparently tunneled over a multipath transport and then 
>> "untunneled" for the final transit to the server. One of the benefits 
>> of QUIC in practice is the ability to potentially bring the transport 
>> implementation "closer" to the application than was traditionally 
>> possible with TCP. One useful signal to applications is the 
>> transport's knowledge of the underlying path state. By transparently 
>> tunneling a QUIC connection over multiple paths this signal is lost. 
>> This has practical implications; the congestion controller used by 
>> the transport cannot proactively act on network switching signals, 
>> and the application cannot proactively adjust its behavior either. As 
>> a trivial example, an ABR streaming video application may detect that 
>> the underlying path has switched and subsequently adjust the 
>> bandwidth. One could imagine partially getting around this problem by 
>> implementing a channel on the UE (i.e. some sort of API) through 
>> which the information about the underlying path state is conveyed to 
>> the QUIC transport which is being tunneled. However, this doesn't 
>> address the loss of information for the server. Additionally, I would 
>> argue that if this UE informational channel exists there is no need 
>> to tunnel QUIC traffic at all -- the QUIC transport being utilized 
>> directly by the application can manage its use of the underlying 
>> network interfaces.
>>
>> Said another way, I think it would be useful to further discuss why 
>> these problems need to be solved in "the network" itself rather than 
>> focusing on making QUIC "natively" capable of optimally navigating 5G 
>> networks.
>>
>>     Best,
>>
>>     Spencer, but not just Spencer
>>
>>     ---------- Forwarded message ---------
>>     From: <internet-drafts@ietf.org <mailto:internet-drafts@ietf.org>>
>>     Date: Sat, May 30, 2020 at 8:46 AM
>>     Subject: New Version Notification for
>>     draft-bonaventure-quic-atsss-overview-00.txt
>>     To: SungHoon Seo <sh.seo@kt.com <mailto:sh.seo@kt.com>>, Olivier
>>     Bonaventure <Olivier.Bonaventure@uclouvain.be
>>     <mailto:Olivier.Bonaventure@uclouvain.be>>, Qing An
>>     <anqing.aq@alibaba-inc.com <mailto:anqing.aq@alibaba-inc.com>>,
>>     Markus Amend <markus.amend@telekom.de
>>     <mailto:markus.amend@telekom.de>>, Andreas Kassler
>>     <andreas.kassler@kau.se <mailto:andreas.kassler@kau.se>>, Spencer
>>     Dawkins <spencerdawkins.ietf@gmail.com
>>     <mailto:spencerdawkins.ietf@gmail.com>>, Mohamed Boucadair
>>     <mohamed.boucadair@orange.com
>>     <mailto:mohamed.boucadair@orange.com>>, Quentin De Coninck
>>     <quentin.deconinck@uclouvain.be
>>     <mailto:quentin.deconinck@uclouvain.be>>, Mirja Kuehlewind
>>     <mirja.kuehlewind@ericsson.com
>>     <mailto:mirja.kuehlewind@ericsson.com>>, Olivier Bonaventure
>>     <olivier.bonaventure@uclouvain.be
>>     <mailto:olivier..bonaventure@uclouvain.be>>, Nicolas Keukeleire
>>     <nicolas.keukeleire@tessares.net
>>     <mailto:nicolas.keukeleire@tessares.net>>, Maxime Piraux
>>     <maxime.piraux@uclouvain.be <mailto:maxime.piraux@uclouvain.be>>
>>
>>
>>
>>
>>     A new version of I-D, draft-bonaventure-quic-atsss-overview-00.txt
>>     has been successfully submitted by Olivier Bonaventure and posted
>>     to the
>>     IETF repository.
>>
>>     Name:  draft-bonaventure-quic-atsss-overview
>>     Revision:       00
>>     Title:          3GPP Access Traffic Steering Switching and
>>     Splitting (ATSSS) - Overview for IETF Participants
>>     Document date:  2020-05-30
>>     Group:          Individual Submission
>>     Pages:          29
>>     URL:
>>     https://www.ietf.org/internet-drafts/draft-bonaventure-quic-atsss-overview-00.txt
>>     Status:
>>     https://datatracker.ietf.org/doc/draft-bonaventure-quic-atsss-overview/
>>     Htmlized:
>>     https://tools.ietf.org/html/draft-bonaventure-quic-atsss-overview-00
>>     Htmlized:
>>     https://datatracker.ietf.org/doc/html/draft-bonaventure-quic-atsss-overview
>>
>>
>>     Abstract:
>>        This document briefly presents the Access Traffic Steering,
>>        Switching, and Splitting (ATSSS) service being specified
>>     within the
>>        3rd Generation Partnership Project (3GPP). The ATSSS service
>>        provides network support for multihomed devices to select a
>>     path for
>>        transmission (steer), move traffic from one path to another
>>     (switch),
>>        or use multiple paths simultaneously (split).  TS 23.501
>>     specifies an
>>        ATSSS architecture for TCP traffic.
>>
>>        This document presents a snap-shot of the ongoing discussion
>>     in the
>>        3GPP to enable ATSSS for non-TCP traffic, based on the use of
>>     QUIC,
>>        and assesses to what extent IETF specifications can be used to
>>     meet
>>        the ATSSS design goals.  Apparent gaps are also documented.
>>
>>
>>
>>
>>     Please note that it may take a couple of minutes from the time of
>>     submission
>>     until the htmlized version and diff are available at
>>     tools.ietf.org <http://tools.ietf.org>.
>>
>>     The IETF Secretariat
>>
>> Matt Joras
>>

--------------C3B0C642C2728766012F67C5
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: 8bit

<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <p>Hi Christian,</p>
    <p><br>
    </p>
    <p>When I understood this right for MPTCP then this can be applied
      end-to-end without any ATSSS interaction. The ATSSS proxy is only
      needed when the end servers do not support MPTCP. When there is
      end-to-end MPQUIC then there doesn't seem to be a need for ATSSS
      as well.<br>
    </p>
    <p><br>
    </p>
    <p>A cellular network often knows something about mobile devices
      even when those devices are not actively using it. A device
      registers with a cellular network and gets an IP address. After
      that the cellular network keeps track of the device location area
      (only an area, not the exact location) even when the device is
      inactive or connected to WiFi. This is done in case some data
      destinated to the device (IP) arrives in the cellular network
      (PGW/UPF). Then the device is paged in the area and the data is
      delivered after the device becomes active. You can of cause turn
      off cellular if you don't want this.</p>
    <p><br>
    </p>
    <p>Regards,</p>
    <p>Roland<br>
    </p>
    <p><br>
    </p>
    <div class="moz-cite-prefix">Am 02.06.2020 um 22:21 schrieb
      Christian Huitema:<br>
    </div>
    <blockquote type="cite"
      cite="mid:72be8104-e738-136f-d05c-285fc49533dc@huitema.net">
      <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
      <p>I think that we should be very clear that the primary goal for
        QUIC multipath design is end-to-end. For example, a device might
        very well have connection with 2 cellular networks, each
        providing their own "advanced" services, and use end-to-end
        multipath to balance traffic between the two networks.</p>
      <p>We also have to consider multipath privacy. In the ATSSS setup,
        the device is supposed to establish a connection with the mobile
        operator's service even in cases when it is only connected to a
        Wi-Fi network. That means disclosing connectivity and activity
        to the operator, just in case a backup through cellular might be
        needed. This is definitely a trade-off that many applications
        will not want to make. End to end multipath alleviates that
        concern, because the network operators only learn of the
        device's activity when the devices actually uses their network.</p>
      <p>-- Christian Huitema<br>
      </p>
      <div class="moz-cite-prefix">On 6/2/2020 12:44 PM, Mirja
        Kuehlewind wrote:<br>
      </div>
      <blockquote type="cite"
        cite="mid:D2BBDD3C-89F7-43BF-B5C3-1EC5E8C69EBE@ericsson.com">
        <meta http-equiv="Content-Type" content="text/html;
          charset=UTF-8">
        <meta name="Generator" content="Microsoft Word 15 (filtered
          medium)">
        <style><!--
/* Font Definitions */
@font-face
	{font-family:"Cambria Math";
	panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
	{font-family:Calibri;
	panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
	{margin:0cm;
	margin-bottom:.0001pt;
	font-size:11.0pt;
	font-family:"Calibri",sans-serif;}
a:link, span.MsoHyperlink
	{mso-style-priority:99;
	color:blue;
	text-decoration:underline;}
span.EmailStyle18
	{mso-style-type:personal-reply;
	font-family:"Calibri",sans-serif;
	color:windowtext;}
..MsoChpDefault
	{mso-style-type:export-only;
	font-size:10.0pt;}
@page WordSection1
	{size:612.0pt 792.0pt;
	margin:72.0pt 72.0pt 72.0pt 72.0pt;}
div.WordSection1
	{page:WordSection1;}
--></style>
        <div class="WordSection1">
          <p class="MsoNormal"><span style="mso-fareast-language:EN-US"
              lang="EN-US">Hi Matt,<o:p></o:p></span></p>
          <p class="MsoNormal"><span style="mso-fareast-language:EN-US"
              lang="EN-US"><o:p> </o:p></span></p>
          <p class="MsoNormal"><span style="mso-fareast-language:EN-US"
              lang="EN-US">thanks for the feedback. Please note that in
              ATSSS the UE can decide to use this multipath service, or
              not, e.g. if the server provides multipath support
              directly (or even do both but not sure how useful that
              is). However, using the network supported service can be
              beneficial for downlink traffic towards the UE as the
              network might have information about network state and
              potential needs to handover. This information can be
              shared with the UE, however, there is no protocol
              mechanism to share this information with the server. I
              believe we briefly mention this in the draft but I guess
              this discussion could definitely be more explicit. <o:p></o:p></span></p>
          <p class="MsoNormal"><span style="mso-fareast-language:EN-US"
              lang="EN-US"><o:p> </o:p></span></p>
          <p class="MsoNormal"><span style="mso-fareast-language:EN-US"
              lang="EN-US">Mirja<o:p></o:p></span></p>
          <p class="MsoNormal"><span style="mso-fareast-language:EN-US"
              lang="EN-US"><o:p> </o:p></span></p>
          <p class="MsoNormal"><span style="mso-fareast-language:EN-US"
              lang="EN-US"><o:p> </o:p></span></p>
          <p class="MsoNormal"><span style="mso-fareast-language:EN-US"><o:p> </o:p></span></p>
          <div style="border:none;border-top:solid #B5C4DF
            1.0pt;padding:3.0pt 0cm 0cm 0cm">
            <p class="MsoNormal" style="margin-left:36.0pt"><b><span
                  style="font-size:12.0pt;color:black">From: </span></b><span
                style="font-size:12.0pt;color:black">QUIC <a
                  class="moz-txt-link-rfc2396E"
                  href="mailto:quic-bounces@ietf.org"
                  moz-do-not-send="true">&lt;quic-bounces@ietf.org&gt;</a>
                on behalf of Matt Joras <a
                  class="moz-txt-link-rfc2396E"
                  href="mailto:matt.joras@gmail.com"
                  moz-do-not-send="true">&lt;matt.joras@gmail.com&gt;</a><br>
                <b>Date: </b>Tuesday, 2. June 2020 at 18:36<br>
                <b>To: </b>Spencer Dawkins at IETF <a
                  class="moz-txt-link-rfc2396E"
                  href="mailto:spencerdawkins.ietf@gmail.com"
                  moz-do-not-send="true">&lt;spencerdawkins.ietf@gmail.com&gt;</a><br>
                <b>Cc: </b>IETF QUIC WG <a
                  class="moz-txt-link-rfc2396E"
                  href="mailto:quic@ietf.org" moz-do-not-send="true">&lt;quic@ietf.org&gt;</a><br>
                <b>Subject: </b>Re: New Version Notification for
                draft-bonaventure-quic-atsss-overview-00.txt<o:p></o:p></span></p>
          </div>
          <div>
            <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
          </div>
          <div>
            <div>
              <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
            </div>
            <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
            <div>
              <div>
                <p class="MsoNormal" style="margin-left:36.0pt">On Tue,
                  Jun 2, 2020 at 4:39 AM Spencer Dawkins at IETF &lt;<a
                    href="mailto:spencerdawkins.ietf@gmail.com"
                    moz-do-not-send="true">spencerdawkins.ietf@gmail.com</a>&gt;
                  wrote:<o:p></o:p></p>
              </div>
              <blockquote style="border:none;border-left:solid #CCCCCC
                1.0pt;padding:0cm 0cm 0cm
                6.0pt;margin-left:4.8pt;margin-right:0cm">
                <div>
                  <p class="MsoNormal" style="margin-left:36.0pt">Dear
                    QUIC, <o:p></o:p></p>
                  <div>
                    <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
                  </div>
                  <div>
                    <p class="MsoNormal" style="margin-left:36.0pt">Several
                      of us submitted this draft (which I emphasize has
                      no special status in IETF, 3GPP, or between IETF
                      and 3GPP).It's being used as one of the inputs to
                      SA2#139E, which is underway now (and that also
                      doesn't give it any special status).. <o:p></o:p></p>
                  </div>
                  <div>
                    <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
                  </div>
                  <div>
                    <p class="MsoNormal" style="margin-left:36.0pt">We
                      are, of course, interested in comments, especially
                      if we've gotten something wrong about QUIC,
                      directions for MP-QUIC, or apparent gaps.<o:p></o:p></p>
                  </div>
                  <div>
                    <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
                  </div>
                  <div>
                    <p class="MsoNormal" style="margin-left:36.0pt">The
                      QUIC chairs have told us this is an appropriate
                      thing to discuss on the QUIC mailing list - we did
                      ask.<o:p></o:p></p>
                  </div>
                </div>
              </blockquote>
              <div>
                <p class="MsoNormal"
style="mso-margin-top-alt:0cm;margin-right:0cm;margin-bottom:12.0pt;margin-left:36.0pt">Thank
                  you very much for producing and sharing this document,
                  I agree it is appropriate for this forum. The diagrams
                  in the document are especially helpful.<o:p></o:p></p>
              </div>
              <div>
                <p class="MsoNormal" style="margin-left:36.0pt">That
                  being said, I have a lot of concerns about the
                  feasibility and usefulness of the schemes in this
                  document. I wonder if this usage of QUIC is in the
                  spirit of many QUIC deployments, or if it potentially
                  hinders an application's usage of QUIC. Correct me if
                  I am wrong, but it seems one of the core tenets of
                  these schemes is that user applications on mobile
                  devices would have their flows to servers
                  transparently tunneled over a multipath transport and
                  then "untunneled" for the final transit to the server.
                  One of the benefits of QUIC in practice is the ability
                  to potentially bring the transport implementation
                  "closer" to the application than was traditionally
                  possible with TCP. One useful signal to applications
                  is the transport's knowledge of the underlying path
                  state. By transparently tunneling a QUIC connection
                  over multiple paths this signal is lost. This has
                  practical implications; the congestion controller used
                  by the transport cannot proactively act on network
                  switching signals, and the application cannot
                  proactively adjust its behavior either. As a trivial
                  example, an ABR streaming video application may detect
                  that the underlying path has switched and subsequently
                  adjust the bandwidth. One could imagine partially
                  getting around this problem by implementing a channel
                  on the UE (i.e. some sort of API) through which the
                  information about the underlying path state is
                  conveyed to the QUIC transport which is being
                  tunneled. However, this doesn't address the loss of
                  information for the server. Additionally, I would
                  argue that if this UE informational channel exists
                  there is no need to tunnel QUIC traffic at all -- the
                  QUIC transport being utilized directly by the
                  application can manage its use of the underlying
                  network interfaces.<o:p></o:p></p>
              </div>
              <div>
                <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
              </div>
              <div>
                <p class="MsoNormal" style="margin-left:36.0pt">Said
                  another way, I think it would be useful to further
                  discuss why these problems need to be solved in "the
                  network" itself rather than focusing on making QUIC
                  "natively" capable of optimally navigating 5G
                  networks.<o:p></o:p></p>
              </div>
              <div>
                <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
              </div>
              <blockquote style="border:none;border-left:solid #CCCCCC
                1.0pt;padding:0cm 0cm 0cm
                6.0pt;margin-left:4.8pt;margin-right:0cm">
                <div>
                  <div>
                    <p class="MsoNormal" style="margin-left:36.0pt"> <o:p></o:p></p>
                  </div>
                  <div>
                    <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
                  </div>
                  <div>
                    <p class="MsoNormal" style="margin-left:36.0pt">Best,<o:p></o:p></p>
                  </div>
                  <div>
                    <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
                  </div>
                  <div>
                    <p class="MsoNormal"
style="mso-margin-top-alt:0cm;margin-right:0cm;margin-bottom:12.0pt;margin-left:36.0pt">Spencer,
                      but not just Spencer<o:p></o:p></p>
                    <div>
                      <div>
                        <p class="MsoNormal" style="margin-left:36.0pt">----------
                          Forwarded message ---------<br>
                          From: &lt;<a
                            href="mailto:internet-drafts@ietf.org"
                            target="_blank" moz-do-not-send="true">internet-drafts@ietf.org</a>&gt;<br>
                          Date: Sat, May 30, 2020 at 8:46 AM<br>
                          Subject: New Version Notification for
                          draft-bonaventure-quic-atsss-overview-00.txt<br>
                          To: SungHoon Seo &lt;<a
                            href="mailto:sh.seo@kt.com" target="_blank"
                            moz-do-not-send="true">sh.seo@kt.com</a>&gt;,
                          Olivier Bonaventure &lt;<a
                            href="mailto:Olivier.Bonaventure@uclouvain.be"
                            target="_blank" moz-do-not-send="true">Olivier.Bonaventure@uclouvain.be</a>&gt;,
                          Qing An &lt;<a
                            href="mailto:anqing.aq@alibaba-inc.com"
                            target="_blank" moz-do-not-send="true">anqing.aq@alibaba-inc.com</a>&gt;,
                          Markus Amend &lt;<a
                            href="mailto:markus.amend@telekom.de"
                            target="_blank" moz-do-not-send="true">markus.amend@telekom.de</a>&gt;,
                          Andreas Kassler &lt;<a
                            href="mailto:andreas.kassler@kau.se"
                            target="_blank" moz-do-not-send="true">andreas.kassler@kau.se</a>&gt;,
                          Spencer Dawkins &lt;<a
                            href="mailto:spencerdawkins.ietf@gmail.com"
                            target="_blank" moz-do-not-send="true">spencerdawkins.ietf@gmail.com</a>&gt;,
                          Mohamed Boucadair &lt;<a
                            href="mailto:mohamed.boucadair@orange.com"
                            target="_blank" moz-do-not-send="true">mohamed.boucadair@orange.com</a>&gt;,
                          Quentin De Coninck &lt;<a
                            href="mailto:quentin.deconinck@uclouvain.be"
                            target="_blank" moz-do-not-send="true">quentin.deconinck@uclouvain.be</a>&gt;,
                          Mirja Kuehlewind &lt;<a
                            href="mailto:mirja.kuehlewind@ericsson.com"
                            target="_blank" moz-do-not-send="true">mirja.kuehlewind@ericsson.com</a>&gt;,
                          Olivier Bonaventure &lt;<a
                            href="mailto:olivier..bonaventure@uclouvain.be"
                            target="_blank" moz-do-not-send="true">olivier.bonaventure@uclouvain.be</a>&gt;,
                          Nicolas Keukeleire &lt;<a
                            href="mailto:nicolas.keukeleire@tessares.net"
                            target="_blank" moz-do-not-send="true">nicolas.keukeleire@tessares.net</a>&gt;,
                          Maxime Piraux &lt;<a
                            href="mailto:maxime.piraux@uclouvain.be"
                            target="_blank" moz-do-not-send="true">maxime.piraux@uclouvain.be</a>&gt;<o:p></o:p></p>
                      </div>
                      <p class="MsoNormal"
style="mso-margin-top-alt:0cm;margin-right:0cm;margin-bottom:12.0pt;margin-left:36.0pt"><br>
                        <br>
                        <br>
                        A new version of I-D,
                        draft-bonaventure-quic-atsss-overview-00.txt<br>
                        has been successfully submitted by Olivier
                        Bonaventure and posted to the<br>
                        IETF repository.<br>
                        <br>
                        Name:         
                         draft-bonaventure-quic-atsss-overview<br>
                        Revision:       00<br>
                        Title:          3GPP Access Traffic Steering
                        Switching and Splitting (ATSSS) - Overview for
                        IETF Participants<br>
                        Document date:  2020-05-30<br>
                        Group:          Individual Submission<br>
                        Pages:          29<br>
                        URL:            <a
href="https://www.ietf.org/internet-drafts/draft-bonaventure-quic-atsss-overview-00.txt"
                          target="_blank" moz-do-not-send="true">
https://www.ietf.org/internet-drafts/draft-bonaventure-quic-atsss-overview-00.txt</a><br>
                        Status:         <a
href="https://datatracker.ietf.org/doc/draft-bonaventure-quic-atsss-overview/"
                          target="_blank" moz-do-not-send="true">https://datatracker.ietf.org/doc/draft-bonaventure-quic-atsss-overview/</a><br>
                        Htmlized:       <a
href="https://tools.ietf.org/html/draft-bonaventure-quic-atsss-overview-00"
                          target="_blank" moz-do-not-send="true">https://tools.ietf.org/html/draft-bonaventure-quic-atsss-overview-00</a><br>
                        Htmlized:       <a
href="https://datatracker.ietf.org/doc/html/draft-bonaventure-quic-atsss-overview"
                          target="_blank" moz-do-not-send="true">https://datatracker.ietf.org/doc/html/draft-bonaventure-quic-atsss-overview</a><br>
                        <br>
                        <br>
                        Abstract:<br>
                           This document briefly presents the Access
                        Traffic Steering,<br>
                           Switching, and Splitting (ATSSS) service
                        being specified within the<br>
                           3rd Generation Partnership Project (3GPP). 
                        The ATSSS service<br>
                           provides network support for multihomed
                        devices to select a path for<br>
                           transmission (steer), move traffic from one
                        path to another (switch),<br>
                           or use multiple paths simultaneously
                        (split).  TS 23.501 specifies an<br>
                           ATSSS architecture for TCP traffic.<br>
                        <br>
                           This document presents a snap-shot of the
                        ongoing discussion in the<br>
                           3GPP to enable ATSSS for non-TCP traffic,
                        based on the use of QUIC,<br>
                           and assesses to what extent IETF
                        specifications can be used to meet<br>
                           the ATSSS design goals.  Apparent gaps are
                        also documented.<br>
                        <br>
                        <br>
                        <br>
                        <br>
                        Please note that it may take a couple of minutes
                        from the time of submission<br>
                        until the htmlized version and diff are
                        available at <a href="http://tools.ietf.org"
                          target="_blank" moz-do-not-send="true">
                          tools.ietf.org</a>.<br>
                        <br>
                        The IETF Secretariat<br>
                        <br>
                        <o:p></o:p></p>
                    </div>
                  </div>
                </div>
              </blockquote>
              <div>
                <p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
              </div>
              <div>
                <p class="MsoNormal" style="margin-left:36.0pt">Matt
                  Joras <o:p></o:p></p>
              </div>
            </div>
          </div>
        </div>
      </blockquote>
    </blockquote>
  </body>
</html>

--------------C3B0C642C2728766012F67C5--


From nobody Tue Jun  2 16:29:34 2020
Return-Path: <huitema@huitema.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 7281E3A1105 for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 16:29:32 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, SPF_HELO_NONE=0.001, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id neOfUtyGODrB for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 16:29:31 -0700 (PDT)
Received: from mx43-out1.antispamcloud.com (mx43-out1.antispamcloud.com [138.201.61.189]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 0DAEE3A1104 for <quic@ietf.org>; Tue,  2 Jun 2020 16:29:31 -0700 (PDT)
Received: from xse310.mail2web.com ([66.113.197.56] helo=xse.mail2web.com) by mx114.antispamcloud.com with esmtp (Exim 4.92) (envelope-from <huitema@huitema.net>) id 1jgGLY-000axq-Mt for quic@ietf.org; Wed, 03 Jun 2020 01:29:29 +0200
Received: from xsmtp22.mail2web.com (unknown [10.100.68.61]) by xse.mail2web.com (Postfix) with ESMTPS id 49c6vK2fw5z6VsQ for <quic@ietf.org>; Tue,  2 Jun 2020 16:00:13 -0700 (PDT)
Received: from [10.5.2.16] (helo=xmail06.myhosting.com) by xsmtp22.mail2web.com with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:256) (Exim 4.92) (envelope-from <huitema@huitema.net>) id 1jgFtF-0000ZJ-8e for quic@ietf.org; Tue, 02 Jun 2020 16:00:13 -0700
Received: (qmail 2545 invoked from network); 2 Jun 2020 23:00:12 -0000
Received: from unknown (HELO [192.168.1.107]) (Authenticated-user:_huitema@huitema.net@[172.58.43.64]) (envelope-sender <huitema@huitema.net>) by xmail06.myhosting.com (qmail-ldap-1.03) with ESMTPA for <quic@ietf.org>; 2 Jun 2020 23:00:12 -0000
To: Roland Zink <roland@zinks.de>, quic@ietf.org
References: <159084638843.27466.7915766554130545967@ietfa.amsl.com> <CAKKJt-eHQtgjc-zuO7vrGZ1Q2c7=3hetOb0FyqnEmbTDu1Uwuw@mail.gmail.com> <CADdTf+iBRLu20OH-WTEmo=e7WZ8Ce5QVP+_LWO09u6LxjCPe2g@mail.gmail.com> <D2BBDD3C-89F7-43BF-B5C3-1EC5E8C69EBE@ericsson.com> <72be8104-e738-136f-d05c-285fc49533dc@huitema.net> <f2d20d2f-d574-e0bb-cd71-35503964c82c@zinks.de>
From: Christian Huitema <huitema@huitema.net>
Autocrypt: addr=huitema@huitema.net; prefer-encrypt=mutual; keydata= mDMEXtavGxYJKwYBBAHaRw8BAQdA1ou9A5MHTP9N3jfsWzlDZ+jPnQkusmc7sfLmWVz1Rmu0 J0NocmlzdGlhbiBIdWl0ZW1hIDxodWl0ZW1hQGh1aXRlbWEubmV0PoiWBBMWCAA+FiEEw3G4 Nwi4QEpAAXUUELAmqKBYtJQFAl7WrxsCGwMFCQlmAYAFCwkIBwIGFQoJCAsCBBYCAwECHgEC F4AACgkQELAmqKBYtJQbMwD/ebj/qnSbthC/5kD5DxZ/Ip0CGJw5QBz/+fJp3R8iAlsBAMjK r2tmyWyJz0CUkVG24WaR5EAJDvgwDv8h22U6QVkAuDgEXtavGxIKKwYBBAGXVQEFAQEHQJoM 6MUAIqpoqdCIiACiEynZf7nlJg2Eu0pXIhbUGONdAwEIB4h+BBgWCAAmFiEEw3G4Nwi4QEpA AXUUELAmqKBYtJQFAl7WrxsCGwwFCQlmAYAACgkQELAmqKBYtJRm2wD7BzeK5gEXSmBcBf0j BYdSaJcXNzx4yPLbP4GnUMAyl2cBAJzcsR4RkwO4dCRqM9CHpVJCwHtbUDJaa55//E0kp+gH
Subject: Re: New Version Notification for draft-bonaventure-quic-atsss-overview-00.txt
Message-ID: <ea81ee5e-22a6-edc0-c2db-d8ce2d05ebe8@huitema.net>
Date: Tue, 2 Jun 2020 16:00:12 -0700
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:68.0) Gecko/20100101 Thunderbird/68.8.0
MIME-Version: 1.0
In-Reply-To: <f2d20d2f-d574-e0bb-cd71-35503964c82c@zinks.de>
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Content-Language: en-US
X-Originating-IP: 66.113.197.56
X-Spampanel-Domain: xsmtpout.mail2web.com
X-Spampanel-Username: 66.113.197.0/24
Authentication-Results: antispamcloud.com; auth=pass smtp.auth=66.113.197.0/24@xsmtpout.mail2web.com
X-Spampanel-Outgoing-Class: unsure
X-Spampanel-Outgoing-Evidence: Combined (0.15)
X-Recommended-Action: accept
X-Filter-ID: Mvzo4OR0dZXEDF/gcnlw0T32zu7lnXf5JHTxKFqDV/OpSDasLI4SayDByyq9LIhVUZbR67CQ7/vm /hHDJU4RXkTNWdUk1Ol2OGx3IfrIJKywOmJyM1qr8uRnWBrbSAGDcnqpk5VeF3xR4kF6iVwRtbgN zB/4Jkrw1eDLcif59fv/UEY9KAbdQsenOUscphKhU7Tmz6iKnkQL9gqsxD347235Nhqq+/HvroPq 8GSPg+60/QPNqXybIny9WGhadIo/d/hBjqsxautjlVXfyJaQKbyme9ldZJ7uNXfg/GfS8fUvP/L5 rCqHDsKZM+xa1iwJX+gRCHfMVnsAk591zk0uilUI+ZL4xWiN8NS6C+dmX6OEdA4u1aThyWrQ/ou2 +v/lmX4Em37yFgrCB6NHRn1g+f3uncIqYSL3lhh5c81YyJqFoLZMmkWsaurVZfvqROaDnDtHb8z5 dpPkEuJ8Snwqla7jUnW3hy14Yji8fo+4xCnSRo4Rcu5Z37rMuDjCny5fE9ykbJ7I9co1MAEE3ruN Xsm8UJsAPvDcVSKtDCYkioPY5Qx4fJOk03R5fJtf/Dv/dkIzS7m4GUpXCY1Y3j3ilUN7TTX3qb0a 8RNcOLCOSd6whjgtKo9vvLdWvMqyXFm2s0ig5FXNj8gTONhiUTNPzzlvekm8sQPjZSMxYCIi5tgv UwPy3x0FYtCNEb10sHyQCLHEvD1OqP6bgZ4L66GcgBg66gs5OuzYxJgw5atIxeNDvjI/CYe5WPy0 +t1RP0azev9O5MZeLDHV7/FDROsLTpxMPnetLBJMh51NiRRoHIC5eUQn1da8r/l5OyX4PvuPmiK7 x42VjdzChZMe6O/DiWiiIzuXMTE3l4bIsk+O50sB+BhWZBrEJLEQJbuZHVw508QV3No+S2msRDep v5w/kkG0v17AmegcpQ0tml/sN9lmMy/o83jVXTcfb9k0nLWblJy7uxV6dw8jzlsaNZe6hynMJcjx DydxsJEju76A7X1QIVydqXpZ6MHhiKws9Iiut28r9wo4SqUIg8Yh9hAM0n3LLzx/F2gT3wl8JQJv Bho=
X-Report-Abuse-To: spam@quarantine11.antispamcloud.com
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/a6I5Xl6ZEWm-OsU4Xk8_Exx-hpA>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 02 Jun 2020 23:29:32 -0000

On 6/2/2020 3:19 PM, Roland Zink wrote:
>
> When I understood this right for MPTCP then this can be applied
> end-to-end without any ATSSS interaction. The ATSSS proxy is only
> needed when the end servers do not support MPTCP. When there is
> end-to-end MPQUIC then there doesn't seem to be a need for ATSSS as wel=
l.
>
Yes, that's pretty much the idea. In fact, we have seen arguments on the
QUIC WG list that the scenarios in which multipath is used for
reliability are already supported with the connection migration features
of QUIC=C2=A0 V1. It is not clear that there is a technical need for ATSS=
S
style proxies in QUIC at all, even with just QUIC V1.
>
> A cellular network often knows something about mobile devices even
> when those devices are not actively using it. A device registers with
> a cellular network and gets an IP address. After that the cellular
> network keeps track of the device location area (only an area, not the
> exact location) even when the device is inactive or connected to WiFi.
> This is done in case some data destinated to the device (IP) arrives
> in the cellular network (PGW/UPF). Then the device is paged in the
> area and the data is delivered after the device becomes active. You
> can of cause turn off cellular if you don't want this.
>
Yes an No. If you do not turn on "airplane mode" or something similar,
the cellular network does indeed track the location of the device -- it
needs to do that if the device is to receive calls. But the cellular
network does not normally receive a copy of the data sent over the Wi-Fi
network. The ATSSS architecture changes that: instead of sending the
data from the device directly to the destination server, the packets
would take a detour through the ATSSS proxy of the cellular network
operator. At a minimum, this enables the cellular network to log the
connection's metadata.

-- Christian Huitema



From nobody Tue Jun  2 17:28:09 2020
Return-Path: <watsonbladd@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 623903A1153 for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 17:28:07 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.098
X-Spam-Level: 
X-Spam-Status: No, score=-2.098 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id HMUz6PfYmwnA for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 17:28:05 -0700 (PDT)
Received: from mail-lf1-x144.google.com (mail-lf1-x144.google.com [IPv6:2a00:1450:4864:20::144]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 5AE3B3A1158 for <quic@ietf.org>; Tue,  2 Jun 2020 17:28:05 -0700 (PDT)
Received: by mail-lf1-x144.google.com with SMTP id c21so191354lfb.3 for <quic@ietf.org>; Tue, 02 Jun 2020 17:28:05 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc:content-transfer-encoding; bh=i/MhagO1CB39m03+Pmzou/FCUIG86HhK6pFD7Gs54TI=; b=avv9l0wUZoMPo/3hT7Zm2U27fFzQCJOdQOYupnL4w9iH5Zw/80z14BdJ5uODkG6mUW DWDzJP4xqAU6XafOJN4um4tVXdaKNydRAgVaxRzDM3zQfgx3BRN1d3fWzUFZw7kCDKYk TCgHkspd5hKAerzhKbSHtPYzaNGiW6goDH6RymmGw1mvy4IJSD+BsMf8FKfFfCpf99k8 UAw+3t3p861vtcDkimE+zBW6IjCDnN3TpytQJNGgHbYLnXtvQeFdeOhj/YjZJXq08sRl bkniPBvDhE3/vzObRCKiNg84Z4q+O14SzwpRlcbHyBXAHCP7I/KItFiAPMtPs/pfyTXN AVxw==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc:content-transfer-encoding; bh=i/MhagO1CB39m03+Pmzou/FCUIG86HhK6pFD7Gs54TI=; b=SJxCwu2sKRgfiGVCdEQ36j6V/yLXInf3TUijf4KnS4AWz0Jr5hHmbrcmGzhuqk/p5Z xrwvcfMETJzRBI1U9xwepSl8+dJO6npLH0mdCzluRuaaCH8jtH0crpu/+PJM+jNJDVsK +U3lIwiphtL6Gv0IEVBUSS6/BglWiXwfHdd+n2rIe84EHOxUcaW7nvdUnN9+7tIXSR1d Ekb8jiEjjIGRwy2lzO0On9u6jmV/MZCOid0C+8qMn0hrYMDVtky0U68vPeF/+l7zqDB+ Ax2+8sZivrcRXyo3SNZsWFCYciFhZhNXbouN37the8wNZa+Y1pNyG0ottmUEPF2fAmV9 k/FQ==
X-Gm-Message-State: AOAM533WrPDoR0Y2iLRZMKqOoq2hTAJOYvXjKM2yFe0unwbJ9TmZu9Vx lZZFKidZgoXYFOgIYtv+sWzh5xTkM+f34yYpEdY=
X-Google-Smtp-Source: ABdhPJy641rbSYR7MzTX7vjVJHLYTNdQff57dNTTU76S+t0JWvhgaSPLhC7Q54PS5Qlz7r1JjVMwSa6xjgmot6Amk2k=
X-Received: by 2002:a19:6cd:: with SMTP id 196mr956998lfg.216.1591144078354; Tue, 02 Jun 2020 17:27:58 -0700 (PDT)
MIME-Version: 1.0
References: <CAM4esxQG=YXThMA6uW3Y=3Qjt5xv8vOt92+FQ7GebcFGFVoH_g@mail.gmail.com> <SN2PR00MB0173119C2F79216A5C0D3750B38E0@SN2PR00MB0173.namprd00.prod.outlook.com> <E1EB91DF-54F9-43E7-90F3-42ED042CB71D@ericsson.com> <CH2PR00MB0726D9D56D3B42831D74A620B68B0@CH2PR00MB0726.namprd00.prod.outlook.com>
In-Reply-To: <CH2PR00MB0726D9D56D3B42831D74A620B68B0@CH2PR00MB0726.namprd00.prod.outlook.com>
From: Watson Ladd <watsonbladd@gmail.com>
Date: Tue, 2 Jun 2020 20:27:46 -0400
Message-ID: <CACsn0c=4JmjLn8Zq66urtkLAL98KcYu81aoXEF8Ct7aJi4hUaA@mail.gmail.com>
Subject: Re: QUIC-LB issues
To: Praveen Balasubramanian <pravb=40microsoft.com@dmarc.ietf.org>
Cc: Mirja Kuehlewind <mirja.kuehlewind=40ericsson.com@dmarc.ietf.org>,  Nick Banks <nibanks=40microsoft.com@dmarc.ietf.org>,  Martin Duke <martin.h.duke@gmail.com>, IETF QUIC WG <quic@ietf.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/XoZsnDSm44LUKf-s2Ql6VbR5w84>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 03 Jun 2020 00:28:07 -0000

On Tue, Jun 2, 2020 at 12:15 PM Praveen Balasubramanian
<pravb=3D40microsoft.com@dmarc.ietf.org> wrote:
>
> >>=E2=80=9CConnection IDs MUST NOT contain any information that can be us=
ed by an external observer (that is, one that does not cooperate with the i=
ssuer) to correlate them with other connection IDs for the same connection.=
=E2=80=9D
>
> That=E2=80=99s quite an onerous requirement and its unclear what is meant=
 by correlation. And since it has nothing to do with interop I don=E2=80=99=
t understand how the MUST NOT can be verified.

Many security related MUSTs have nothing to do with interop. As for
what correlation of connections means it means that given a set of
connection IDs and the packets it should be impossible to group those
into connections with greater then chance success.

>
>
>
> From: QUIC <quic-bounces@ietf.org> On Behalf Of Mirja Kuehlewind
> Sent: Tuesday, June 2, 2020 3:34 AM
> To: Nick Banks <nibanks=3D40microsoft.com@dmarc.ietf.org>; Martin Duke <m=
artin.h.duke@gmail.com>; IETF QUIC WG <quic@ietf.org>
> Subject: [EXTERNAL] Re: QUIC-LB issues
>
>
>
> Hi Martin,
>
>
>
> I can=E2=80=99t provide you any feedback on implementation or deployment =
plans but given these algorithms are used with one domain, I would assume t=
hat some plaintext algorithms would be used anyway even if not documented i=
n this draft. Therefore I would rather like see all the different approache=
s documented and discussed here. And there is already a clear warning in th=
e security consideration section.
>
>
>
> Not sure about the need for a new transport parameter. I guess given the =
server selects the CID there could always be a risk that they are somehow l=
inkable. There is no way for the client to check that. So if a client reall=
y, really wants to ensure that there is no linkability, it probably general=
ly has to disable migration entirely.
>
>
>
> However, I would also like to note that the current transport draft says =
the following:
>
> =E2=80=9CConnection IDs MUST NOT contain any information that can be used=
 by an external observer (that is, one that does not cooperate with the iss=
uer) to correlate them with other connection IDs for the same connection.=
=E2=80=9D
>
>
>
> Mirja
>
>
>
>
>
>
>
> From: QUIC <quic-bounces@ietf.org> on behalf of Nick Banks <nibanks=3D40m=
icrosoft.com@dmarc.ietf.org>
> Date: Thursday, 28. May 2020 at 16:41
> To: Martin Duke <martin.h.duke@gmail.com>, IETF QUIC WG <quic@ietf.org>
> Subject: RE: QUIC-LB issues
>
>
>
> Hello Martin & WG,
>
>
>
> For Azure, we=E2=80=99re in the process of building a load balancing sche=
me on top of the Plaintext CID algorithm, so we=E2=80=99d definitely like t=
o keep it in the document. The complicated nature of some of our scenarios =
(multiple containers per-DIP, all behind a single LB; with all components g=
enerally independent) currently prevents us from practically supporting any=
thing more than the most basic algorithm. Additionally, whether it=E2=80=99=
s adopted by the WG or something Azure specific, we will likely require an =
in-band communication to communicate the Server ID to all the routing compo=
nents involved.
>
>
>
> We may eventually adopt one of the ciphertext algorithms in the future, b=
ut I doubt we=E2=80=99d ever add support for the Obfuscated CID algorithm.
>
>
>
> On the topic of issue #8, we disagree that the possible information expos=
ure when using the Plaintext CID algorithm is enough reason to declare the =
CIDs as =E2=80=9Cguessable=E2=80=9D. So we=E2=80=99d be against recommendin=
g/requiring that migration be disabled when using the plaintext algorithm. =
Additionally, in our opinion, if migration is not supported/allowed by the =
server, then there isn=E2=80=99t really any reason to do anything more than=
 IP tuple load balancing at all. The whole point of QUIC-LB is to provide f=
or load balancing solutions that survive tuple changes.
>
>
>
> Thanks,
>
> - Nick
>
>
>
> From: QUIC <quic-bounces@ietf.org> On Behalf Of Martin Duke
> Sent: Tuesday, May 26, 2020 11:20 AM
> To: IETF QUIC WG <quic@ietf.org>
> Subject: QUIC-LB issues
>
>
>
> Hello all,
>
>
>
> There hasn't been much activity in QUIC-LB lately. There is at least one =
issue that could use some broader community input.
>
>
>
> https://github.com/quicwg/load-balancers/issues/8
>
>
>
> The core issue is that there are 4 algorithms to encode the server ID, 2 =
ciphertext and 2 plaintext. Of the two plaintext algorithms, one is blindin=
gly obvious (just encode it a known field) and the other applies quite a bi=
t of obfuscation. The obfuscated version has undergone some informal analys=
is vs. brute force attacks, but I doubt crypto experts would be impressed.
>
>
>
> The QUIC WG reflex has generally been to encrypt everything possible. The=
re are two reasons theseplaintext algorithms exist:
>
> 1) As section 2.2 describes, there is no such thing as perfect protection=
 from linkability, so even full encryption is only moving things along a co=
ntinuum. It is not nearly as bright a line as, say, https vs. http.
>
> 2) What little feedback we've gotten from people that make layer 4 cloud =
load balancers has been pretty negative about adding encryption to their im=
plementations. I can probably get F5 to adopt it, but other than that the d=
eployment story is not solid on the load balancer side. There is still much=
 work to do here on outreach to the big cloud providers.
>
>
>
> Possibly, an open source implementation (I'm working on it...) might assu=
age their fears. An implementation that implements both LB-side encrypted a=
lgorithms is less than 80 lines of C, not including openssl or the separate=
 code to configure the LB properly.
>
>
>
> The design team had consensus to include all of these options, with some =
intending to support them in their products. Beyond that, Martin T is the o=
nly person I'm aware of to have engaged on the PT options, and he was prett=
y negative about them. What is the sense of the group as a whole?
>
> 1. Should this document include the Plaintext CID (PCID) algorithm?
>
> 2. Should this document include the Obfuscated CID (OCID) algorithm?
>
>
>
> *****
>
>
>
> On a related note: one issue with the plaintext/obfuscated algorithms is =
that they are chosen by the server but the consequences of linkability are =
borne by the client. The current draft says that the non-obfuscated algorit=
hm SHOULD be accompanied by a disable_active_migration parameter; that's po=
tentially an overloaded codepoint.
>
>
>
> https://github.com/quicwg/load-balancers/issues/16
>
>
>
> An alternative would be to create a new transport parameter that would al=
low the server to explicitly say what it is doing. Clients could then decid=
e whether or not to migrate with their eyes wide open. On the other hand, t=
his leaks some information about the encoding into the network.
>
>
>
> 3. Assuming we have at least one unencrypted algorithm, is this server TP=
 a good idea or not?



--=20
"Man is born free, but everywhere he is in chains".
--Rousseau.


From nobody Tue Jun  2 23:31:39 2020
Return-Path: <olivier.bonaventure@uclouvain.be>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C7D863A07B3 for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 23:31:36 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.1
X-Spam-Level: 
X-Spam-Status: No, score=-2.1 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, MSGID_FROM_MTA_HEADER=0.001, RCVD_IN_MSPIKE_H2=-0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=uclouvain.be
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id oh9a8qTM9Fsz for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 23:31:34 -0700 (PDT)
Received: from EUR02-HE1-obe.outbound.protection.outlook.com (mail-eopbgr10104.outbound.protection.outlook.com [40.107.1.104]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 305553A07A5 for <quic@ietf.org>; Tue,  2 Jun 2020 23:31:33 -0700 (PDT)
ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=hEGnQX1EuxN3/VQBhGwwfo7rz0kBYlPn1FgqXbdbxIOojL2yLlsT6vNShpD6aMogDT7pNe+kBkLVq9UnMBCwyVjJPpHaf8SX9XgEVAkUjP3Lf4bhm9fgROG7ic7kR2KXx+7cBuS1YTLW4zunTwVeK2/FJcKdQEAtBuGg0Z3Vj9xafEp+xR/L4IFdPU0AcllW3ZVTpvx9r1mQg7biznn2UGRrqQBPlrADY02ePGDHwSHDUUfSr1jkyyO0RDdWXGqM3p1gSVJJTOWX8tGat1AWMJRKAJN4n3LCc4XjOjZOo5TLyWWtL8jpXnfW4fsR0RHCY0fzW5/yBifAfxaB72qo6A==
ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com;  s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=nE7x+kBxs1Gs8ptg45tnfhjMhhOqaOqZNBONYo8ZLCI=; b=Z/IAkMETfvo1fiySSku93swMbTMBY3m4GSAKsjlZf9fg7C4pfU9SaFVUh27VhiIRiQ/tIX4pRmMphACwoaJ5xd3ZnslxsJrpkoOc8KCxRSMdR4AaX5IL9IU/tabppLXBvWvd83CCrJJHB1geGP3yItBhqa9oByayFECb8G2c/JicU0htx/sjoWXGdt7BVZdnnoflQdtGH8iAo3BrOKFTip+dnt9jnZFsKjyQOekg2+vPbOc5R7eZgyj8ADMlkmoQRzog026gTrZCHck1wFjgTvuYWN4rxej4GkghkXL48DdbMVaP++/ARKbQ5JKcvmIKOLrVTFvnGscXvkBdqKNOrg==
ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=uclouvain.be; dmarc=pass action=none header.from=uclouvain.be; dkim=pass header.d=uclouvain.be; arc=none
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=uclouvain.be; s=selector2; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=nE7x+kBxs1Gs8ptg45tnfhjMhhOqaOqZNBONYo8ZLCI=; b=PrZImo39Oh49gBVxUZGsYcJA7OQGbdnQIU+QH9XAyoPTiWkUvGtupssRszSd7JuQJIJHMPDgSYSWxYua5OvmoGH2y2Dxka+B/9WON21UeEhT7Sde5OGsvSi4sBVEmLBcLDoRDp9gc+F6OphpsdcHN38PrKe0K4VGgAyVoxrOaFM=
Authentication-Results: ietf.org; dkim=none (message not signed) header.d=none;ietf.org; dmarc=none action=none header.from=uclouvain.be;
Received: from AM7PR03MB6642.eurprd03.prod.outlook.com (2603:10a6:20b:1bf::6) by AM7PR03MB6546.eurprd03.prod.outlook.com (2603:10a6:20b:1bf::13) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.3045.19; Wed, 3 Jun 2020 06:31:29 +0000
Received: from AM7PR03MB6642.eurprd03.prod.outlook.com ([fe80::fcc2:e19d:eec3:15e6]) by AM7PR03MB6642.eurprd03.prod.outlook.com ([fe80::fcc2:e19d:eec3:15e6%7]) with mapi id 15.20.3066.018; Wed, 3 Jun 2020 06:31:29 +0000
Reply-To: Olivier.Bonaventure@uclouvain.be
Subject: Re: New Version Notification for draft-bonaventure-quic-atsss-overview-00.txt
To: Matt Joras <matt.joras@gmail.com>, Spencer Dawkins at IETF <spencerdawkins.ietf@gmail.com>
Cc: IETF QUIC WG <quic@ietf.org>
References: <159084638843.27466.7915766554130545967@ietfa.amsl.com> <CAKKJt-eHQtgjc-zuO7vrGZ1Q2c7=3hetOb0FyqnEmbTDu1Uwuw@mail.gmail.com> <CADdTf+iBRLu20OH-WTEmo=e7WZ8Ce5QVP+_LWO09u6LxjCPe2g@mail.gmail.com>
From: Olivier Bonaventure <Olivier.Bonaventure@uclouvain.be>
Message-ID: <0ecd2f86-3472-5242-eb41-6eaefb640027@uclouvain.be>
Date: Wed, 3 Jun 2020 08:31:28 +0200
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.14; rv:68.0) Gecko/20100101 Thunderbird/68.8.1
In-Reply-To: <CADdTf+iBRLu20OH-WTEmo=e7WZ8Ce5QVP+_LWO09u6LxjCPe2g@mail.gmail.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: fr-classic
Content-Transfer-Encoding: 7bit
X-ClientProxiedBy: AM4PR0902CA0016.eurprd09.prod.outlook.com (2603:10a6:200:9b::26) To AM7PR03MB6642.eurprd03.prod.outlook.com (2603:10a6:20b:1bf::6)
MIME-Version: 1.0
X-MS-Exchange-MessageSentRepresentingType: 1
Received: from mbpobo-2.local (2a02:2788:484:b4f:85fe:72b8:b48d:9955) by AM4PR0902CA0016.eurprd09.prod.outlook.com (2603:10a6:200:9b::26) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.3066.18 via Frontend Transport; Wed, 3 Jun 2020 06:31:29 +0000
X-Originating-IP: [2a02:2788:484:b4f:85fe:72b8:b48d:9955]
X-MS-PublicTrafficType: Email
X-MS-Office365-Filtering-Correlation-Id: 37157a23-0d89-481d-d739-08d80787bfd4
X-MS-TrafficTypeDiagnostic: AM7PR03MB6546:
X-Microsoft-Antispam-PRVS: <AM7PR03MB6546D4D46B91419936D5BCF986880@AM7PR03MB6546.eurprd03.prod.outlook.com>
X-MS-Oob-TLC-OOBClassifiers: OLM:10000;
X-Forefront-PRVS: 04238CD941
X-MS-Exchange-SenderADCheck: 1
X-Microsoft-Antispam: BCL:0;
X-Microsoft-Antispam-Message-Info: bchO4q+Syf3vmpjClqG8c41831JZ2J+jbPXlFp5bD/8+01Bc8PUGNrJcxUoIua+cgnOL6MfFi1Q7aWtpH/OJFHJQ52nHKxtXn5xykWPx0a6uZ6Y6vltnPSeIi4TDkGOSCMr2Rj6WWkgSHVQhSGVIpN1J2Nh11sUGVaSqrGy1GcJzDlLvbT7xZkCdoBPeRWrHuXxwyS/u06bnfAJhUfqvqc5Z9EQScKm6tUn9mLLrzHReeb21Y8tKfc8YgCEqcc5cUYUFA2ncKHgV5fmR7fuqtZir2a6EXynnuNnNGIntDkAqx/palPNn2qH6JHttlzZsX5FuaAWlrLQ/EUWnVsVALGyz5oErjXq6jUqkF6XHPnGCDPECpW091qpOCl4+ZK8N
X-Forefront-Antispam-Report: CIP:255.255.255.255; CTRY:; LANG:en; SCL:1; SRV:;  IPV:NLI; SFV:NSPM; H:AM7PR03MB6642.eurprd03.prod.outlook.com; PTR:; CAT:NONE;  SFTY:; SFS:(4636009)(376002)(346002)(366004)(39850400004)(396003)(136003)(16526019)(86362001)(66946007)(52116002)(186003)(66556008)(66476007)(83380400001)(36756003)(6506007)(31696002)(6486002)(4326008)(6512007)(5660300002)(316002)(786003)(3450700001)(2616005)(478600001)(8676002)(2906002)(8936002)(110136005)(31686004)(43740500002); DIR:OUT; SFP:1102; 
X-MS-Exchange-AntiSpam-MessageData: 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
X-OriginatorOrg: uclouvain.be
X-MS-Exchange-CrossTenant-Network-Message-Id: 37157a23-0d89-481d-d739-08d80787bfd4
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 03 Jun 2020 06:31:29.5253 (UTC)
X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted
X-MS-Exchange-CrossTenant-Id: 7ab090d4-fa2e-4ecf-bc7c-4127b4d582ec
X-MS-Exchange-CrossTenant-MailboxType: HOSTED
X-MS-Exchange-CrossTenant-UserPrincipalName: lRFF4SuJVnJ+N56qvmANY4yUept0Rh2zjqbABCB7j5hwm572kL7aqirYn2jL8to24fADW+F/bA6HoCdXeAZpkGfprAYlqOG1TAJBHz6lVGPRn0/Q/zx8eG6ezOVPuJih
X-MS-Exchange-Transport-CrossTenantHeadersStamped: AM7PR03MB6546
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/tCiddwlA1BLkYe09BCRS-_Qq7aE>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 03 Jun 2020 06:31:37 -0000

Matt,
> 
> Thank you very much for producing and sharing this document, I agree it 
> is appropriate for this forum. The diagrams in the document are 
> especially helpful.
> 
> That being said, I have a lot of concerns about the feasibility and 
> usefulness of the schemes in this document. I wonder if this usage of 
> QUIC is in the spirit of many QUIC deployments, or if it potentially 
> hinders an application's usage of QUIC. 

This ATSSS solution should be considered as another application for QUIC 
that leverages the flexibility of QUIC. With the datagram extension, it 
becomes possible to carry unreliable datagrams over a 
congestion-controlled QUIC connection to the ATSSS function that is 
provided by the network operator. This could be considered as a kind of 
operator-provided "VPN" service that allows unmodified applications 
running on enduser devices to efficiently use the available paths 
without having to modify these applications.

> Correct me if I am wrong, but it 
> seems one of the core tenets of these schemes is that user applications 
> on mobile devices would have their flows to servers transparently 
> tunneled over a multipath transport and then "untunneled" for the final 
> transit to the server. One of the benefits of QUIC in practice is the 
> ability to potentially bring the transport implementation "closer" to 
> the application than was traditionally possible with TCP. One useful 
> signal to applications is the transport's knowledge of the underlying 
> path state. By transparently tunneling a QUIC connection over multiple 
> paths this signal is lost. This has practical implications; the 
> congestion controller used by the transport cannot proactively act on 
> network switching signals, and the application cannot proactively adjust 
> its behavior either. As a trivial example, an ABR streaming video 
> application may detect that the underlying path has switched and 
> subsequently adjust the bandwidth. One could imagine partially getting 
> around this problem by implementing a channel on the UE (i.e. some sort 
> of API) through which the information about the underlying path state is 
> conveyed to the QUIC transport which is being tunneled. However, this 
> doesn't address the loss of information for the server. Additionally, I 
> would argue that if this UE informational channel exists there is no 
> need to tunnel QUIC traffic at all -- the QUIC transport being utilized 
> directly by the application can manage its use of the underlying network 
> interfaces.
> 
> Said another way, I think it would be useful to further discuss why 
> these problems need to be solved in "the network" itself rather than 
> focusing on making QUIC "natively" capable of optimally navigating 5G 
> networks.

Solving these problems in the network allows to support existing and 
unmodifed applications, which is a key requirement from a deployment 
viewpoint. A longer term approach could be to expose more information 
from the underlying network to the upper layer so that new applications 
can make informed and possibly better choices.


Olivier


From nobody Tue Jun  2 23:41:28 2020
Return-Path: <olivier.bonaventure@uclouvain.be>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 92D8B3A0770 for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 23:41:26 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.1
X-Spam-Level: 
X-Spam-Status: No, score=-2.1 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, MSGID_FROM_MTA_HEADER=0.001, RCVD_IN_MSPIKE_H2=-0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=uclouvain.be
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ckyfMOBiOKUC for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 23:41:24 -0700 (PDT)
Received: from EUR03-DB5-obe.outbound.protection.outlook.com (mail-eopbgr40093.outbound.protection.outlook.com [40.107.4.93]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 95D783A0B98 for <quic@ietf.org>; Tue,  2 Jun 2020 23:41:24 -0700 (PDT)
ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=OQkDWSdHKT7F6rB+jRfEJTKQceD/elmLbZlMiTozhQnBFXKB2gTgzY7rDAE2Qo7agOm/2YY12YQixtDl0TOc5D1sSOfxjztmogUg47DgPI5XHL4RUiuVY4SBgKMqP+CzehWv6/oUOrWOUMMwCPovuvL83txEbh6pVirZTJoIJYHmvGmKpcnSG8/VhoSp1hi13dunhtqwRUO5fHylnEw5y231igmbkdsEd4tN7QhXUgBxt+MJ0atIl12KjfB7inC3RtKxXOiNkY1VJ1Eyt8BaMgRyA83L3WUDwbCsq/NOCN7+iAixa+qYGk8m50Pxmt9JUWZhpQ24U+dWy+KG5aiQuQ==
ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com;  s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=VVPovMVh50814CIlT8yVHg1H+REUqdtMi9PDFhA6O6o=; b=bJYUsL4MqHhkFQ9NzU7qqZXhkiPJYBqqinV4aSmNH1yK44RPLxUnTA5DjrZYLrfP4N79xPkau/eBpMPaikYthNGO7h1ideIpvH+/n7R9+/27vsc7HLrrGk3NSffmdF3cQYMef0JtsWMIlLyRlX9Ipj9cmOJ1lSgEt3uIpMm6i/DCB5ERu8Y5wLXpLoIRD4TDzkaPT9SFQtj7BNMeQCgmtoQMVZQF+wcdVMKcvgp2+zPd/no/gKuo8OH+ag0ApneeXVibCksdH0iobS8kdDNYegDPlMbMr4nXJYokRTHsGWr5GJzCWK01Ui7N/6fLyNBLNrRvkthTkEB9Kb+B2rXfvQ==
ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=uclouvain.be; dmarc=pass action=none header.from=uclouvain.be; dkim=pass header.d=uclouvain.be; arc=none
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=uclouvain.be; s=selector2; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=VVPovMVh50814CIlT8yVHg1H+REUqdtMi9PDFhA6O6o=; b=oth1W5GfOtn/vRGYiQCqDFT4eIj/y1Vk51AkbS5lj7wjMy1ENQTopvCv2ON1a+2gfdrUWYq/PyMWduo+cTANvGdEwEcNG5cxNXj3Z8HHjyRoBAW2jpdk56s0SNT8OSzRo0j0b2ltLcdOIZ6/uob7ey4OOMvm6BJnd8jNbt5PgOw=
Authentication-Results: ietf.org; dkim=none (message not signed) header.d=none;ietf.org; dmarc=none action=none header.from=uclouvain.be;
Received: from AM7PR03MB6642.eurprd03.prod.outlook.com (2603:10a6:20b:1bf::6) by AM7PR03MB6232.eurprd03.prod.outlook.com (2603:10a6:20b:13b::24) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.3045.19; Wed, 3 Jun 2020 06:41:22 +0000
Received: from AM7PR03MB6642.eurprd03.prod.outlook.com ([fe80::fcc2:e19d:eec3:15e6]) by AM7PR03MB6642.eurprd03.prod.outlook.com ([fe80::fcc2:e19d:eec3:15e6%7]) with mapi id 15.20.3066.018; Wed, 3 Jun 2020 06:41:21 +0000
Reply-To: Olivier.Bonaventure@uclouvain.be
Subject: Re: New Version Notification for draft-bonaventure-quic-atsss-overview-00.txt
To: Christian Huitema <huitema@huitema.net>, Mirja Kuehlewind <mirja.kuehlewind=40ericsson.com@dmarc.ietf.org>, Matt Joras <matt.joras@gmail.com>, Spencer Dawkins at IETF <spencerdawkins.ietf@gmail.com>
Cc: IETF QUIC WG <quic@ietf.org>
References: <159084638843.27466.7915766554130545967@ietfa.amsl.com> <CAKKJt-eHQtgjc-zuO7vrGZ1Q2c7=3hetOb0FyqnEmbTDu1Uwuw@mail.gmail.com> <CADdTf+iBRLu20OH-WTEmo=e7WZ8Ce5QVP+_LWO09u6LxjCPe2g@mail.gmail.com> <D2BBDD3C-89F7-43BF-B5C3-1EC5E8C69EBE@ericsson.com> <72be8104-e738-136f-d05c-285fc49533dc@huitema.net>
From: Olivier Bonaventure <Olivier.Bonaventure@uclouvain.be>
Message-ID: <e74e1342-4f54-679b-00f3-a2e2dc24c9d0@uclouvain.be>
Date: Wed, 3 Jun 2020 08:41:20 +0200
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.14; rv:68.0) Gecko/20100101 Thunderbird/68.8.1
In-Reply-To: <72be8104-e738-136f-d05c-285fc49533dc@huitema.net>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: fr-classic
Content-Transfer-Encoding: 7bit
X-ClientProxiedBy: AM4PR0101CA0050.eurprd01.prod.exchangelabs.com (2603:10a6:200:41::18) To AM7PR03MB6642.eurprd03.prod.outlook.com (2603:10a6:20b:1bf::6)
MIME-Version: 1.0
X-MS-Exchange-MessageSentRepresentingType: 1
Received: from mbpobo-2.local (2a02:2788:484:b4f:85fe:72b8:b48d:9955) by AM4PR0101CA0050.eurprd01.prod.exchangelabs.com (2603:10a6:200:41::18) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.3045.19 via Frontend Transport; Wed, 3 Jun 2020 06:41:21 +0000
X-Originating-IP: [2a02:2788:484:b4f:85fe:72b8:b48d:9955]
X-MS-PublicTrafficType: Email
X-MS-Office365-Filtering-Correlation-Id: 0f436f22-9076-4cd2-5345-08d8078920eb
X-MS-TrafficTypeDiagnostic: AM7PR03MB6232:
X-Microsoft-Antispam-PRVS: <AM7PR03MB62322AADEACDD63125AAD8FA86880@AM7PR03MB6232.eurprd03.prod.outlook.com>
X-MS-Oob-TLC-OOBClassifiers: OLM:9508;
X-Forefront-PRVS: 04238CD941
X-MS-Exchange-SenderADCheck: 1
X-Microsoft-Antispam: BCL:0;
X-Microsoft-Antispam-Message-Info: /C7wPzfFjhlyXIzTReqCOx0Gs3OjyTsVN5NfHZrDtxk1NpMgD+yXf/7HLuQi0q3sJyxaXq1jWfhwoBZDR4Ahkq6wijqU3P92tTA3QZSzVroF5xijxzgMCyDDm3lq/l/xj13ig3xn/brbl4pOXm67UU1iGaT7JrNHP3ceSdKJjpZBi6lICt/IgUEV6EmGODJ39LKUInJLoGzb8LakDg9PA0cRecZbD869pPWW3UCUxRYXRXuq69PL8M+vaAjQdv46RhHjpuES2Xd9RTashsqRlNtUAbP/ExgvpkJHN6b31gURDfD1yaq6/j9CxXidYvf9aSpeiYlJO9ZG+IYAUG31EoigdM0BEpl/u5PIwIv9TCjCFEgsMaW2nGFGhwyTXLcB
X-Forefront-Antispam-Report: CIP:255.255.255.255; CTRY:; LANG:en; SCL:1; SRV:;  IPV:NLI; SFV:NSPM; H:AM7PR03MB6642.eurprd03.prod.outlook.com; PTR:; CAT:NONE;  SFTY:; SFS:(4636009)(376002)(39850400004)(136003)(396003)(346002)(366004)(2906002)(16526019)(186003)(6506007)(86362001)(8676002)(316002)(110136005)(3450700001)(8936002)(2616005)(31696002)(478600001)(36756003)(6486002)(786003)(66556008)(83380400001)(31686004)(4326008)(66574014)(66476007)(52116002)(66946007)(5660300002)(6512007)(43740500002); DIR:OUT; SFP:1102; 
X-MS-Exchange-AntiSpam-MessageData: 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
X-OriginatorOrg: uclouvain.be
X-MS-Exchange-CrossTenant-Network-Message-Id: 0f436f22-9076-4cd2-5345-08d8078920eb
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 03 Jun 2020 06:41:21.8496 (UTC)
X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted
X-MS-Exchange-CrossTenant-Id: 7ab090d4-fa2e-4ecf-bc7c-4127b4d582ec
X-MS-Exchange-CrossTenant-MailboxType: HOSTED
X-MS-Exchange-CrossTenant-UserPrincipalName: qJPKSwtQkf/zxoxZ2FMQyseFq1SV3YnZK5KxwR9N50co7U7Kc6WhVyZiapmg1tsmtFuf5b4QOTgMpPgsU92sOlJHN2K8ITrQUhpIaLGqNLb/NKKuhFJlbDWWadbdBsmQ
X-MS-Exchange-Transport-CrossTenantHeadersStamped: AM7PR03MB6232
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/5UiXQsZqQzOHeRVqzQ75f0o4p_w>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 03 Jun 2020 06:41:27 -0000

Christian,

> I think that we should be very clear that the primary goal for QUIC 
> multipath design is end-to-end. For example, a device might very well 
> have connection with 2 cellular networks, each providing their own 
> "advanced" services, and use end-to-end multipath to balance traffic 
> between the two networks.

This end-to-end approach would indeed be possible with a multipath QUIC 
design. Note that the ATSSS service includes rules that determine which 
applications benefit from the service. It does not imply that all 
applications will use it.

AFAIK, many existing QUIC deployments are not really end-to-end 
deployments. Often, clients interact with an edge server that 
caches/proxies on behalf of origin servers.
> 
> We also have to consider multipath privacy. In the ATSSS setup, the 
> device is supposed to establish a connection with the mobile operator's 
> service even in cases when it is only connected to a Wi-Fi network. That 
> means disclosing connectivity and activity to the operator, just in case 
> a backup through cellular might be needed. This is definitely a 
> trade-off that many applications will not want to make. End to end 
> multipath alleviates that concern, because the network operators only 
> learn of the device's activity when the devices actually uses their network.

There are different ways to consider the privacy discussion. As an 
employee, my company might require me to always use a VPN service to 
access the Internet. That's a common policy in many companies. The ATSSS 
service can be considered as a kind of "VPN" service that network 
operators provide to their customers. As an enduser, I have a 
contractual relationship with my network provider and this contract can 
include strong privacy protections. Using the ATSSS service of a 
provider that I trust and needs to obey to specific laws could be better 
from a privacy viewpoint.

Olivier


From nobody Tue Jun  2 23:48:58 2020
Return-Path: <lars@eggert.org>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 5B4CB3A0C1B for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 23:48:57 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.099
X-Spam-Level: 
X-Spam-Status: No, score=-2.099 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=unavailable autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=eggert.org
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 7CGkFt-YeuKe for <quic@ietfa.amsl.com>; Tue,  2 Jun 2020 23:48:56 -0700 (PDT)
Received: from mail.eggert.org (mail.eggert.org [91.190.195.94]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 5556C3A0C3C for <quic@ietf.org>; Tue,  2 Jun 2020 23:48:56 -0700 (PDT)
Received: from [IPv6:2a00:ac00:0:35:2546:2eeb:f496:7429] (unknown [IPv6:2a00:ac00:0:35:2546:2eeb:f496:7429]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.eggert.org (Postfix) with ESMTPSA id 4018361CFF0; Wed,  3 Jun 2020 09:48:46 +0300 (EEST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=eggert.org; s=dkim; t=1591166926; bh=Ae2SO3OliSPLq9xcfpe0sFGgs1HSEuz4vcI0znGWpYY=; h=From:Subject:Date:In-Reply-To:Cc:To:References; b=e/i2tLmZvO2Dev0kCXFYDOLPTnfypHMZZkVLGeOv7OytlkvuBx9YE0zLHkTc8S2NM V1l0i1/p3IB6xSMgg3t1rdcZssOiAx8qKpfyVT2varwQVE4QDy0qQotB4vTTXLd2mD iitDxlDAf+5joDaggMYA6pptCXNW4gnR3o98siGc=
From: Lars Eggert <lars@eggert.org>
Message-Id: <9F66D3B3-DBD6-44CE-80EA-0F46F3D1DB9A@eggert.org>
Content-Type: multipart/signed; boundary="Apple-Mail=_201694B0-2654-4D5A-A319-ABBF7E0D353B"; protocol="application/pgp-signature"; micalg=pgp-sha512
Mime-Version: 1.0 (Mac OS X Mail 13.4 \(3608.80.23.2.2\))
Subject: Re: New Version Notification for draft-bonaventure-quic-atsss-overview-00.txt
Date: Wed, 3 Jun 2020 09:48:45 +0300
In-Reply-To: <e74e1342-4f54-679b-00f3-a2e2dc24c9d0@uclouvain.be>
Cc: Christian Huitema <huitema@huitema.net>, Mirja Kuehlewind <mirja.kuehlewind=40ericsson.com@dmarc.ietf.org>, Matt Joras <matt.joras@gmail.com>, Spencer Dawkins at IETF <spencerdawkins.ietf@gmail.com>, IETF QUIC WG <quic@ietf.org>
To: Olivier Bonaventure <Olivier.Bonaventure@uclouvain.be>
References: <159084638843.27466.7915766554130545967@ietfa.amsl.com> <CAKKJt-eHQtgjc-zuO7vrGZ1Q2c7=3hetOb0FyqnEmbTDu1Uwuw@mail.gmail.com> <CADdTf+iBRLu20OH-WTEmo=e7WZ8Ce5QVP+_LWO09u6LxjCPe2g@mail.gmail.com> <D2BBDD3C-89F7-43BF-B5C3-1EC5E8C69EBE@ericsson.com> <72be8104-e738-136f-d05c-285fc49533dc@huitema.net> <e74e1342-4f54-679b-00f3-a2e2dc24c9d0@uclouvain.be>
X-MailScanner-ID: 4018361CFF0.A70D7
X-MailScanner: Found to be clean
X-MailScanner-From: lars@eggert.org
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/oAAv_dHrGGEngbbIYAKxZDBd_7E>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 03 Jun 2020 06:48:57 -0000

--Apple-Mail=_201694B0-2654-4D5A-A319-ABBF7E0D353B
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=us-ascii

Hi,

(hat off)

On 2020-6-3, at 9:41, Olivier Bonaventure =
<olivier.bonaventure@uclouvain.be> wrote:
> As an enduser, I have a contractual relationship with my network =
provider and this contract can include strong privacy protections. Using =
the ATSSS service of a provider that I trust and needs to obey to =
specific laws could be better from a privacy viewpoint.

so that's certainly debatable.

Apart from the privacy aspect though, tunneling all traffic through an =
operator proxy that would otherwise not be in the path certainly can =
have performance and management aspects. It's probably going to add =
delay due to route stretch and possibly queueing, and can limit =
throughput due to processing overheads or overload.

Lars

--Apple-Mail=_201694B0-2654-4D5A-A319-ABBF7E0D353B
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
	filename=signature.asc
Content-Type: application/pgp-signature;
	name=signature.asc
Content-Description: Message signed with OpenPGP

-----BEGIN PGP SIGNATURE-----
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=Vi5b
-----END PGP SIGNATURE-----

--Apple-Mail=_201694B0-2654-4D5A-A319-ABBF7E0D353B--


From nobody Wed Jun  3 00:11:29 2020
Return-Path: <huitema@huitema.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id F0CD83A0C9A for <quic@ietfa.amsl.com>; Wed,  3 Jun 2020 00:11:26 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, SPF_HELO_NONE=0.001, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id kt3vtm7jMQmc for <quic@ietfa.amsl.com>; Wed,  3 Jun 2020 00:11:25 -0700 (PDT)
Received: from mx36-out10.antispamcloud.com (mx36-out10.antispamcloud.com [209.126.121.30]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id EF2AA3A0C99 for <quic@ietf.org>; Wed,  3 Jun 2020 00:11:24 -0700 (PDT)
Received: from xse352.mail2web.com ([66.113.197.98] helo=xse.mail2web.com) by mx114.antispamcloud.com with esmtp (Exim 4.92) (envelope-from <huitema@huitema.net>) id 1jgNYX-000wpB-TL for quic@ietf.org; Wed, 03 Jun 2020 09:11:24 +0200
Received: from xsmtp22.mail2web.com (unknown [10.100.68.61]) by xse.mail2web.com (Postfix) with ESMTPS id 49cK5r4xhRz2459 for <quic@ietf.org>; Tue,  2 Jun 2020 23:40:00 -0700 (PDT)
Received: from [10.5.2.31] (helo=xmail09.myhosting.com) by xsmtp22.mail2web.com with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:256) (Exim 4.92) (envelope-from <huitema@huitema.net>) id 1jgN4C-0002EU-JL for quic@ietf.org; Tue, 02 Jun 2020 23:40:00 -0700
Received: (qmail 5685 invoked from network); 3 Jun 2020 06:40:00 -0000
Received: from unknown (HELO [192.168.1.107]) (Authenticated-user:_huitema@huitema.net@[172.58.43.64]) (envelope-sender <huitema@huitema.net>) by xmail09.myhosting.com (qmail-ldap-1.03) with ESMTPA for <quic@ietf.org>; 3 Jun 2020 06:40:00 -0000
Subject: Re: New Version Notification for draft-bonaventure-quic-atsss-overview-00.txt
To: Olivier.Bonaventure@uclouvain.be, Matt Joras <matt.joras@gmail.com>, Spencer Dawkins at IETF <spencerdawkins.ietf@gmail.com>
Cc: IETF QUIC WG <quic@ietf.org>
References: <159084638843.27466.7915766554130545967@ietfa.amsl.com> <CAKKJt-eHQtgjc-zuO7vrGZ1Q2c7=3hetOb0FyqnEmbTDu1Uwuw@mail.gmail.com> <CADdTf+iBRLu20OH-WTEmo=e7WZ8Ce5QVP+_LWO09u6LxjCPe2g@mail.gmail.com> <0ecd2f86-3472-5242-eb41-6eaefb640027@uclouvain.be>
From: Christian Huitema <huitema@huitema.net>
Autocrypt: addr=huitema@huitema.net; prefer-encrypt=mutual; keydata= mDMEXtavGxYJKwYBBAHaRw8BAQdA1ou9A5MHTP9N3jfsWzlDZ+jPnQkusmc7sfLmWVz1Rmu0 J0NocmlzdGlhbiBIdWl0ZW1hIDxodWl0ZW1hQGh1aXRlbWEubmV0PoiWBBMWCAA+FiEEw3G4 Nwi4QEpAAXUUELAmqKBYtJQFAl7WrxsCGwMFCQlmAYAFCwkIBwIGFQoJCAsCBBYCAwECHgEC F4AACgkQELAmqKBYtJQbMwD/ebj/qnSbthC/5kD5DxZ/Ip0CGJw5QBz/+fJp3R8iAlsBAMjK r2tmyWyJz0CUkVG24WaR5EAJDvgwDv8h22U6QVkAuDgEXtavGxIKKwYBBAGXVQEFAQEHQJoM 6MUAIqpoqdCIiACiEynZf7nlJg2Eu0pXIhbUGONdAwEIB4h+BBgWCAAmFiEEw3G4Nwi4QEpA AXUUELAmqKBYtJQFAl7WrxsCGwwFCQlmAYAACgkQELAmqKBYtJRm2wD7BzeK5gEXSmBcBf0j BYdSaJcXNzx4yPLbP4GnUMAyl2cBAJzcsR4RkwO4dCRqM9CHpVJCwHtbUDJaa55//E0kp+gH
Message-ID: <fe1eb46c-a46c-9c63-adc9-b0ec529e0445@huitema.net>
Date: Tue, 2 Jun 2020 23:40:00 -0700
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:68.0) Gecko/20100101 Thunderbird/68.8.0
MIME-Version: 1.0
In-Reply-To: <0ecd2f86-3472-5242-eb41-6eaefb640027@uclouvain.be>
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
Content-Language: en-US
X-Originating-IP: 66.113.197.98
X-Spampanel-Domain: xsmtpout.mail2web.com
X-Spampanel-Username: 66.113.197.0/24
Authentication-Results: antispamcloud.com; auth=pass smtp.auth=66.113.197.0/24@xsmtpout.mail2web.com
X-Spampanel-Outgoing-Class: unsure
X-Spampanel-Outgoing-Evidence: Combined (0.61)
X-Recommended-Action: accept
X-Filter-ID: Mvzo4OR0dZXEDF/gcnlw0T32zu7lnXf5JHTxKFqDV/OpSDasLI4SayDByyq9LIhVjjp2pemLoXVc byihDHYtV0TNWdUk1Ol2OGx3IfrIJKywOmJyM1qr8uRnWBrbSAGDcnqpk5VeF3xR4kF6iVwRtbgN zB/4Jkrw1eDLcif59fsidhM4KiFHSd/l/AXKMxJDU7Tmz6iKnkQL9gqsxD347235Nhqq+/HvroPq 8GSPg+60/QPNqXybIny9WGhadIo/d/hBjqsxautjlVXfyJaQKbyme9ldZJ7uNXfg/GfS8fUvP/L5 rCqHDsKZM+xa1iwJX+gRCHfMVnsAk591zk0uilUI+ZL4xWiN8NS6C+dmX6OEdA4u1aThyWrQ/ou2 +v/lmX4Em37yFgrCB6NHRn1g+f3uncIqYSL3lhh5c81YyJqFoLZMmkWsaurVZfvqROaDnDtHb8z5 dpPkEuJ8Snwqla7jUnW3hy14Yji8fo+4xCnSRo4Rcu5Z37rMuDjCny5fE9ykbJ7I9co1MAEE3ruN Xsm8UJsAPvDcVSKtDCYkioPY5Qx4fJOk03R5fJtf/Dv/dkIzS7m4GUpXCY1Y3j3ilUN7TTX3qb0a 8RNcOLCOSd6whjgtKo9vvLdWvMqyXFm2ImZ/p/QpFwFCjgyYAUAoxpn/Ul2tLdLVS8NDXH/y+KEv UwPy3x0FYtCNEb10sHyQCLHEvD1OqP6bgZ4L66GcgBg66gs5OuzYxJgw5atIxeNDvjI/CYe5WPy0 +t1RP0azQ/SjolqPkJdZxBtoUF0DjpxMPnetLBJMh51NiRRoHIDtW7wUsiIiAjdeaTwfdIHXmiK7 x42VjdzChZMe6O/DiWiiIzuXMTE3l4bIsk+O50sKl4hpKwzZLQKRiWAPqIkM08QV3No+S2msRDep v5w/kkG0v17AmegcpQ0tml/sN9lmMy/o83jVXTcfb9k0nLWblJy7uxV6dw8jzlsaNZe6hynMJcjx DydxsJEju76A7X1QIVydqXpZ6MHhiKws9Iiut28r9wo4SqUIg8Yh9hAM0n3LLzx/F2gT3wl8JQJv Bho=
X-Report-Abuse-To: spam@quarantine11.antispamcloud.com
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/dcBgGsEd7r60BQ4uJ19rQptjaPM>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 03 Jun 2020 07:11:27 -0000

On 6/2/2020 11:31 PM, Olivier Bonaventure wrote:
> Solving these problems in the network allows to support existing and
> unmodifed applications, which is a key requirement from a deployment
> viewpoint. A longer term approach could be to expose more information
> from the underlying network to the upper layer so that new
> applications can make informed and possibly better choices. 

I really wonder whether that argument applies to QUIC in 2020. Existing
and unmodified applications use TCP today, not QUIC.

QUIC capable applications have path migration capabilities built-in from
the start, which covers most of the "reliability" use cases of
multipath, without insert yet another middle-box in the setup.

-- Christian Huitema


From nobody Wed Jun  3 00:26:11 2020
Return-Path: <olivier.bonaventure@uclouvain.be>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C2EE53A0CB2 for <quic@ietfa.amsl.com>; Wed,  3 Jun 2020 00:26:09 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.1
X-Spam-Level: 
X-Spam-Status: No, score=-2.1 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, MSGID_FROM_MTA_HEADER=0.001, RCVD_IN_MSPIKE_H2=-0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=uclouvain.be
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id N42_iPzyGBzz for <quic@ietfa.amsl.com>; Wed,  3 Jun 2020 00:26:08 -0700 (PDT)
Received: from EUR01-HE1-obe.outbound.protection.outlook.com (mail-eopbgr130114.outbound.protection.outlook.com [40.107.13.114]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id D14623A0CB1 for <quic@ietf.org>; Wed,  3 Jun 2020 00:26:07 -0700 (PDT)
ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=eihtBqVgwKKe5ymgz4WLTapn4l2PhXYYjLz///YIvbiPmLXJhkkC/hJy6zQ4Za74SX5zRxmeKwHz5vXEtT/4Il7nv2E55HlBxvhJkNqPcwWuhWiaXx608HZoU/8QwrFYsuic7YoLJjAIyKS48NVLgZk0LRB6UuPETSjCuDaiKT/suiiieqXboegX/WGTJafXfrVg36A/6tJGqeAg1igufgGu/O1h4FyUbHXYh6YRSsfT6XEqCZ077aaJ3C2gVb8cZEEyYgd6IMCqUoqCaHgJoJWiGHC4VS6BNb6D6HlwG5j52VvVaObXd4FoFBqGrI1knFTFusBcai27sF/AmsnMDA==
ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com;  s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=fe27PsVoH7L8NauT/61mLqW9Dyrt/jM1uicD5ZooIAs=; b=bE3v2YoBwdI+LQyGr2LYS40jaRDHu588+hxy40hmltwzJ0h7CoCFQrHAp9SCRJMAnivP1rLo4RRrzATBJVPH65lhLOnct7WH7u2xu0uYFwyIP9dIN+Pk26xMPV9ajenM8fBI50XD5ZyaXMmQa2SVFJfjfHtuR1vXHGxAqgqyx2LThRVpSQiLwIIg77GcKo4vnDWM7WuXAWqKnnlMmcdfO8UQRRR/nfBmNq7SJ4TNq8T+yJixhQu5zWf25XvaRZwpjKCW8N3c0stLgtgc16rG3sVGpjsrKlIqJoSyGp9PcHRdqQFIF4k/9lZEso8DqKzPlTsQqUtwyPQx6PwPU9j73Q==
ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=uclouvain.be; dmarc=pass action=none header.from=uclouvain.be; dkim=pass header.d=uclouvain.be; arc=none
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=uclouvain.be; s=selector2; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=fe27PsVoH7L8NauT/61mLqW9Dyrt/jM1uicD5ZooIAs=; b=PBp+VzHCuRf3sjH8b4rQhwnUDsHNMITxF1Wt+jRO5csQAhtSnn8Y73m4cvWqtHuT0XS/GzfgaJxW/hnqNq07BbR/gqQTNdkqFdg7Q33SM/FXQE2ZZHda/MPybRlov4NiPediVQPDn5nDPGMdbTHIx4BqS5XpakrPz6DSYqODpIE=
Authentication-Results: ietf.org; dkim=none (message not signed) header.d=none;ietf.org; dmarc=none action=none header.from=uclouvain.be;
Received: from AM7PR03MB6642.eurprd03.prod.outlook.com (2603:10a6:20b:1bf::6) by AM7PR03MB6215.eurprd03.prod.outlook.com (2603:10a6:20b:13d::21) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.3066.18; Wed, 3 Jun 2020 07:26:03 +0000
Received: from AM7PR03MB6642.eurprd03.prod.outlook.com ([fe80::fcc2:e19d:eec3:15e6]) by AM7PR03MB6642.eurprd03.prod.outlook.com ([fe80::fcc2:e19d:eec3:15e6%7]) with mapi id 15.20.3066.018; Wed, 3 Jun 2020 07:26:03 +0000
Reply-To: Olivier.Bonaventure@uclouvain.be
Subject: Re: New Version Notification for draft-bonaventure-quic-atsss-overview-00.txt
To: Christian Huitema <huitema@huitema.net>, Matt Joras <matt.joras@gmail.com>, Spencer Dawkins at IETF <spencerdawkins.ietf@gmail.com>
Cc: IETF QUIC WG <quic@ietf.org>
References: <159084638843.27466.7915766554130545967@ietfa.amsl.com> <CAKKJt-eHQtgjc-zuO7vrGZ1Q2c7=3hetOb0FyqnEmbTDu1Uwuw@mail.gmail.com> <CADdTf+iBRLu20OH-WTEmo=e7WZ8Ce5QVP+_LWO09u6LxjCPe2g@mail.gmail.com> <0ecd2f86-3472-5242-eb41-6eaefb640027@uclouvain.be> <fe1eb46c-a46c-9c63-adc9-b0ec529e0445@huitema.net>
From: Olivier Bonaventure <Olivier.Bonaventure@uclouvain.be>
Message-ID: <1fa3eabf-451c-722c-1c56-b5f93a72e4d4@uclouvain.be>
Date: Wed, 3 Jun 2020 09:26:02 +0200
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.14; rv:68.0) Gecko/20100101 Thunderbird/68.8.1
In-Reply-To: <fe1eb46c-a46c-9c63-adc9-b0ec529e0445@huitema.net>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: fr-classic
Content-Transfer-Encoding: 7bit
X-ClientProxiedBy: AM0PR10CA0021.EURPRD10.PROD.OUTLOOK.COM (2603:10a6:208:17c::31) To AM7PR03MB6642.eurprd03.prod.outlook.com (2603:10a6:20b:1bf::6)
MIME-Version: 1.0
X-MS-Exchange-MessageSentRepresentingType: 1
Received: from mbpobo-2.local (2a02:2788:484:b4f:dd71:cdaf:a4fb:47ac) by AM0PR10CA0021.EURPRD10.PROD.OUTLOOK.COM (2603:10a6:208:17c::31) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.3066.18 via Frontend Transport; Wed, 3 Jun 2020 07:26:03 +0000
X-Originating-IP: [2a02:2788:484:b4f:dd71:cdaf:a4fb:47ac]
X-MS-PublicTrafficType: Email
X-MS-Office365-Filtering-Correlation-Id: 4292cc02-2cfd-42e5-efdd-08d8078f5f2a
X-MS-TrafficTypeDiagnostic: AM7PR03MB6215:
X-Microsoft-Antispam-PRVS: <AM7PR03MB62158ED64558FD94649B72F586880@AM7PR03MB6215.eurprd03.prod.outlook.com>
X-MS-Oob-TLC-OOBClassifiers: OLM:9508;
X-Forefront-PRVS: 04238CD941
X-MS-Exchange-SenderADCheck: 1
X-Microsoft-Antispam: BCL:0;
X-Microsoft-Antispam-Message-Info: RY0pm+IS1nlX/KTL8D9Xnf6i5F2IF8QW8qvpZ9nOcCJgaO6f0bLsMbJuViVx3wd42fXDPAOyOMIKAPE/lJKHOGfsmJsndEHswtWbXIHfquqcUwOEPh62dlXCaGVRFqJewGGBvn86zaKAt96PPWfybZmq9GiK3hBQmClRJqz7DR4UxwGGwompg0CuZdUUIgcJpOwYp71cs98uieB6qntHc2dPUS9+3xXvzueeH6EKwI2weqfmcBiRgoLhGMsCmVvP0xNIAroGnfUothl3Mls5MoeguC9cP+RNlN2qbJuMCpWu7s2wotz7+G5YdI/6s42vUpOLtE3uteFwF+50c6VE67ZjNFUGKcKWEU/jEwBW/lHCAtdzsgSzoWEYhmEUtdaA
X-Forefront-Antispam-Report: CIP:255.255.255.255; CTRY:; LANG:en; SCL:1; SRV:;  IPV:NLI; SFV:NSPM; H:AM7PR03MB6642.eurprd03.prod.outlook.com; PTR:; CAT:NONE;  SFTY:; SFS:(4636009)(346002)(39860400002)(366004)(376002)(136003)(396003)(478600001)(31696002)(316002)(110136005)(66476007)(786003)(66556008)(66946007)(6506007)(5660300002)(86362001)(52116002)(2616005)(6512007)(8936002)(16526019)(4326008)(36756003)(8676002)(186003)(2906002)(6486002)(3450700001)(31686004)(43740500002); DIR:OUT; SFP:1102; 
X-MS-Exchange-AntiSpam-MessageData: 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
X-OriginatorOrg: uclouvain.be
X-MS-Exchange-CrossTenant-Network-Message-Id: 4292cc02-2cfd-42e5-efdd-08d8078f5f2a
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 03 Jun 2020 07:26:03.5225 (UTC)
X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted
X-MS-Exchange-CrossTenant-Id: 7ab090d4-fa2e-4ecf-bc7c-4127b4d582ec
X-MS-Exchange-CrossTenant-MailboxType: HOSTED
X-MS-Exchange-CrossTenant-UserPrincipalName: 7DIo0nznCP2+v/wgbajuLgBp3IgJnGjI9zS784RqEkUwY7XDu/YoAw7ewCFUy7hEwsJ1wSnwteAOM+Re3Rs73CBp6Z0mySIq8XAAol1HCtdd0nUHPZgOh5aZPnmzEH1O
X-MS-Exchange-Transport-CrossTenantHeadersStamped: AM7PR03MB6215
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/rw5sLVU2uzjl8MyfuAbFA6JiRcg>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 03 Jun 2020 07:26:10 -0000

Christian,

>> Solving these problems in the network allows to support existing and
>> unmodifed applications, which is a key requirement from a deployment
>> viewpoint. A longer term approach could be to expose more information
>> from the underlying network to the upper layer so that new
>> applications can make informed and possibly better choices.
> 
> I really wonder whether that argument applies to QUIC in 2020. Existing
> and unmodified applications use TCP today, not QUIC.

One of the use cases for ATSSSv2 is to simply carry Ethernet frames over 
the cellular and non-cellular networks.

> QUIC capable applications have path migration capabilities built-in from
> the start, which covers most of the "reliability" use cases of
> multipath, without insert yet another middle-box in the setup.

Path migration is a nice feature of QUICv1. I'd be interested in seeing 
measurements that compare QUICv1 path migration during handovers with 
MPTCP's approach to handover allows to use both paths during the 
handover. Experience with MPTCP shows that MPTCP's approach works well. 
If some has data on path migration with QUICv1 applications that do 
handovers, I'd be interested in looking at them.


Olivier


From nobody Wed Jun  3 10:41:58 2020
Return-Path: <ted.ietf@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 9AE3F3A0E06 for <quic@ietfa.amsl.com>; Wed,  3 Jun 2020 10:41:56 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_NONE=-0.0001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id dy48KKGKqGiw for <quic@ietfa.amsl.com>; Wed,  3 Jun 2020 10:41:55 -0700 (PDT)
Received: from mail-oi1-x22b.google.com (mail-oi1-x22b.google.com [IPv6:2607:f8b0:4864:20::22b]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 0C9063A0DEF for <quic@ietf.org>; Wed,  3 Jun 2020 10:41:55 -0700 (PDT)
Received: by mail-oi1-x22b.google.com with SMTP id a137so2572262oii.3 for <quic@ietf.org>; Wed, 03 Jun 2020 10:41:55 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=0BWwMobUT8LZegK9KT84+1ZMZGy69gLiQzfF/X2Ljv4=; b=U/i+1MCDrqZ/iU99HbZejWxwa7b/pFx+zjKpGAypuDUuWsmLDJ/4YehMwu8hk1qs18 hhLuC9GiRGLgpoIdaUBMmMHmP4Zxt749GkllXG5tMhw8yw8Xhc76x/SXUQsju/js2/F5 bo2EhJkKqlz0AdC2N+ldEWOZPO7z8SnAOYhSrjCK9++GHBMiQ2GO3TAhN3jmpsFSCex4 lQvqyeUjDsDGyU8ODJDyjVGb3mz760A4svwe+gZpqgoX0DmmQ+qTA8e2dPD9G/dM7ZB4 VHXKEZYhsvaAVVPJGRkSpE0ddT5/ou+YOsU/vLXAeeURoZsimg+98W/6yg6bAM6j2bzp lu0A==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=0BWwMobUT8LZegK9KT84+1ZMZGy69gLiQzfF/X2Ljv4=; b=eIk8WsuwlXldEaGk/rZK7i36p0eF51MsxEK7HQ6DxW6Cx0nP3+JZl5LITA8Sm45p4f rEzFhM1IbPHdjQTrURZUXZFaEgz4sFhbdMIZ3oprQQpIh3AOX3rhckry2Oqk4yo01Mw6 6I6eWu06jmR78+/WAkHwy6/9q9YK1ybk0TkItIJARRN9U4Jo/rxJTcrT1ekySRyG1zxP 6QNHzeAUNS6sL4OzIxtrlJ8QKmiSYFrAwX+UUi+gZJxT47xOKmLjyFLo6lFu+9FA08GC tvOcP18HkOnXNGbk9S4m1Jm+k6iOA1s6WfM2vZE/g1YsFZ2k07J+dMVPWEFVlCI/n6Qk 236Q==
X-Gm-Message-State: AOAM5338JCcaax4Xp+49pEgYNjVO+sr7gZJVG+dR4hcWo9oETGisT31D HoITpeFw1VL5nGedMUO2A4OIM0EspfAn3ThqBOk=
X-Google-Smtp-Source: ABdhPJz5tFKJ7rw2lsMwOgpF27riIZAYm0DAPAdBTn9clNIXBSXW5RiaYS1NV6cjh5jvKdQMjBQhx2H+LHw6UMcsz/4=
X-Received: by 2002:a54:4795:: with SMTP id o21mr577818oic.74.1591206114376; Wed, 03 Jun 2020 10:41:54 -0700 (PDT)
MIME-Version: 1.0
References: <159084638843.27466.7915766554130545967@ietfa.amsl.com> <CAKKJt-eHQtgjc-zuO7vrGZ1Q2c7=3hetOb0FyqnEmbTDu1Uwuw@mail.gmail.com> <CADdTf+iBRLu20OH-WTEmo=e7WZ8Ce5QVP+_LWO09u6LxjCPe2g@mail.gmail.com> <D2BBDD3C-89F7-43BF-B5C3-1EC5E8C69EBE@ericsson.com> <72be8104-e738-136f-d05c-285fc49533dc@huitema.net> <e74e1342-4f54-679b-00f3-a2e2dc24c9d0@uclouvain.be> <9F66D3B3-DBD6-44CE-80EA-0F46F3D1DB9A@eggert.org>
In-Reply-To: <9F66D3B3-DBD6-44CE-80EA-0F46F3D1DB9A@eggert.org>
From: Ted Hardie <ted.ietf@gmail.com>
Date: Wed, 3 Jun 2020 10:41:28 -0700
Message-ID: <CA+9kkMDhaxZunXxyiHdUyCB=90WQthYDCvajGXL3TEyLcmr5vA@mail.gmail.com>
Subject: Re: New Version Notification for draft-bonaventure-quic-atsss-overview-00.txt
To: Lars Eggert <lars@eggert.org>
Cc: Olivier Bonaventure <Olivier.Bonaventure@uclouvain.be>,  Mirja Kuehlewind <mirja.kuehlewind=40ericsson.com@dmarc.ietf.org>,  IETF QUIC WG <quic@ietf.org>, Christian Huitema <huitema@huitema.net>,  Spencer Dawkins at IETF <spencerdawkins.ietf@gmail.com>, Matt Joras <matt.joras@gmail.com>
Content-Type: multipart/alternative; boundary="00000000000036410b05a7318bee"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/vYXM2LSRoU3Cnbd6D0ltiX4hFiY>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 03 Jun 2020 17:41:57 -0000

--00000000000036410b05a7318bee
Content-Type: text/plain; charset="UTF-8"

Just to amplify this a bit:

On Tue, Jun 2, 2020 at 11:49 PM Lars Eggert <lars@eggert.org> wrote:

>
> Apart from the privacy aspect though, tunneling all traffic through an
> operator proxy that would otherwise not be in the path certainly can have
> performance and management aspects. It's probably going to add delay due to
> route stretch and possibly queueing, and can limit throughput due to
> processing overheads or overload.
>
>
We've been through a lot of this before with various flavors of IP-layer
mobility, and the Home Agent approaches definitely showed all the issues
Lars cites above. There have been attempts to manage that with hierarchies,
fast hand-over rather than tunneling, and even new architectures like LISP
and RINA.   I don't think we want to re-learn all of those lessons from
scratch.  At the very least, we want to start with the understanding that
if ATSSS goes with this design that other tunneling protocols running on
top of it will be very difficult to debug and very hard to keep within
usable latencies.  Another comment mentioned VPNs, which are common and may
become more so.  If you have a user that has to deal with interaction
between the triangle routing to an operator proxy and the triangle routing
to a VPN end point, it will not be quick.

Traffic steering without full knowledge can manage some local
optimization,  and I don't want to denigrate that.  But those local
optimums do not necessarily result in better performance even for single
flows when the endpoints are not clear to steering mechanism.  If that is
priority, sticking with endpoint-based approaches rather than network
steering may be both cheaper and better.

regards,

Ted Hardie

--00000000000036410b05a7318bee
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div>Just to amplify this a bit:<br></div><div dir=3D"ltr"=
><br></div><div dir=3D"ltr">On Tue, Jun 2, 2020 at 11:49 PM Lars Eggert &lt=
;<a href=3D"mailto:lars@eggert.org">lars@eggert.org</a>&gt; wrote:<br></div=
><div class=3D"gmail_quote"><blockquote class=3D"gmail_quote" style=3D"marg=
in:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1e=
x"><br>
Apart from the privacy aspect though, tunneling all traffic through an oper=
ator proxy that would otherwise not be in the path certainly can have perfo=
rmance and management aspects. It&#39;s probably going to add delay due to =
route stretch and possibly queueing, and can limit throughput due to proces=
sing overheads or overload.<br>
<br></blockquote><div><br></div><div>We&#39;ve been through a lot of this b=
efore with various flavors of IP-layer mobility, and the Home Agent approac=
hes definitely showed all the issues Lars cites above. There have been atte=
mpts to manage that with hierarchies, fast hand-over rather than tunneling,=
 and even new architectures like LISP and RINA.=C2=A0=C2=A0 I don&#39;t thi=
nk we want to re-learn all of those lessons from scratch.=C2=A0 At the very=
 least, we want to start with the understanding that if ATSSS goes with thi=
s design that other tunneling protocols running on top of it will be very d=
ifficult to debug and very hard to keep within usable latencies.=C2=A0 Anot=
her comment mentioned VPNs, which are common and may become more so.=C2=A0 =
If you have a user that has to deal with interaction between the triangle r=
outing to an operator proxy and the triangle routing to a VPN end point, it=
 will not be quick. <br></div><div><br></div><div>Traffic steering without =
full knowledge can manage some local optimization,=C2=A0 and I don&#39;t wa=
nt to denigrate that.=C2=A0 But those local optimums do not necessarily res=
ult in better performance even for single flows when the endpoints are not =
clear to steering mechanism.=C2=A0 If that is priority, sticking with endpo=
int-based approaches rather than network steering may be both cheaper and b=
etter.</div><div><br></div><div>regards,</div><div><br></div><div>Ted Hardi=
e<br></div><div><br></div><div><br></div></div></div>

--00000000000036410b05a7318bee--


From nobody Wed Jun  3 11:40:36 2020
Return-Path: <gorry@erg.abdn.ac.uk>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 9AC603A0DE3 for <quic@ietfa.amsl.com>; Wed,  3 Jun 2020 11:40:35 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.897
X-Spam-Level: 
X-Spam-Status: No, score=-1.897 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, SPF_HELO_NONE=0.001, SPF_NONE=0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id QX5GnoRdmIsH for <quic@ietfa.amsl.com>; Wed,  3 Jun 2020 11:40:33 -0700 (PDT)
Received: from pegasus.erg.abdn.ac.uk (pegasus.erg.abdn.ac.uk [IPv6:2001:630:42:150::2]) by ietfa.amsl.com (Postfix) with ESMTP id 648EC3A0DDC for <quic@ietf.org>; Wed,  3 Jun 2020 11:40:33 -0700 (PDT)
Received: from GF-MacBook-Pro.lan (fgrpf.plus.com [212.159.18.54]) by pegasus.erg.abdn.ac.uk (Postfix) with ESMTPSA id F20ED1B00140 for <quic@ietf.org>; Wed,  3 Jun 2020 19:40:28 +0100 (BST)
From: Gorry Fairhurst <gorry@erg.abdn.ac.uk>
Subject: Copies of my slides showing QUIC v. TCP perfomance over asymmetric paths (#3529)
To: quic@ietf.org
Message-ID: <cdd99ef5-5b50-4697-558b-c5f9aa2c0d3c@erg.abdn.ac.uk>
Date: Wed, 3 Jun 2020 19:40:27 +0100
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.13; rv:68.0) Gecko/20100101 Thunderbird/68.8.1
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 7bit
Content-Language: en-GB
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/SY_P3pD8pSjL0q7dUq4jK2AGqkU>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 03 Jun 2020 18:40:36 -0000

For those interested, I presented my slides today in PANRG on the impact 
of QUIC ACKs on performance and the impact of choosing to ACK every 
other ACK-eliciting packet, and the benefits of using an ACK 1:10 - such 
as that in Chromium.

This relates to #3529, if you missed this and want to see read the 
slides they are here:

https://datatracker.ietf.org/meeting/interim-2020-panrg-01/materials/slides-interim-2020-panrg-01-sessa-etosat-update

Gorry



From nobody Fri Jun  5 15:26:57 2020
Return-Path: <lucaspardue.24.7@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 7BCB23A0EBC for <quic@ietfa.amsl.com>; Fri,  5 Jun 2020 15:26:56 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.848
X-Spam-Level: 
X-Spam-Status: No, score=-1.848 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_ENVFROM_END_DIGIT=0.25, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 4BMBKZX_Bh2T for <quic@ietfa.amsl.com>; Fri,  5 Jun 2020 15:26:55 -0700 (PDT)
Received: from mail-wm1-x32d.google.com (mail-wm1-x32d.google.com [IPv6:2a00:1450:4864:20::32d]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 413613A0E95 for <quic@ietf.org>; Fri,  5 Jun 2020 15:26:55 -0700 (PDT)
Received: by mail-wm1-x32d.google.com with SMTP id r9so9718170wmh.2 for <quic@ietf.org>; Fri, 05 Jun 2020 15:26:55 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=tyfC9W5jJkGhyYK4sUQzTfVQRgBhL88WGe9TZgGQqyQ=; b=gahDKCT9nmYf6Sxpzf1r5tzCfoDK9dbNslkxemeeyRm1gGhxQeBXcB8Ylg3x44DJeX HYoNNwutSt1cNCAR2s0EIpgt1p3/iB83aSQL595/oEbjTHjPWJKQFOUuHePtjBltJFKv ZMHuSPJg53tmnAKBqma8mpmjI91C0ZncB22w/VANu2pUuErDzTErY1xdKlTxjMd+5uLO ngOmRzUO5onlMSx8di2j6Qcb5U+uvKAVHhYIyP6qH2kmfFIZjhky+zh9K1cytQU9yfAA H8u23nCUh2y44XDQy5sNLhplcv6dr+W8B9uh3f2DzCu8DNy+UykAHkJLI5CBl8tWFExk 2Anw==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=tyfC9W5jJkGhyYK4sUQzTfVQRgBhL88WGe9TZgGQqyQ=; b=hb5Xm21Xf9cue+c8aZxtaGc+h03tC3QACwWPkoM00gf9HYsnCZZ1b2Ga2rlGHQwhEE x3W1bJPj2cGUT+zxn6iHlGHFjROxtQgYc9cSrpFYpDMtUm5kcbaaHveLsqtn2E0CuM6g KYsehFiOaHldKswWx2u4f3bKr+3IDOGIeqbJ08Ml4UOUD4T7VzLp7J2DNTqCm/+RpKVu a08UjCj7WisIw41g+wUZrrKNFU/tarU9HXNN46VhbV1V+DeeCLcFslsLhrngaWPiHb1e u4tAoX10bw2e4LSiU6TFe7FyGusDPNHIPBQDpkiUWZUUwnX2aRo328+OY5ci27HJCW+0 R3XQ==
X-Gm-Message-State: AOAM532iZ7MJ01gOMrBPVh+PWJGLPeN5BXcvuzHlQ8vaTrXHlkhhNVA9 4qSKYavP6CK43jMtsGeoXDm7MxWJ9w45lZDcvCYhXevuJOM=
X-Google-Smtp-Source: ABdhPJxIdPHw8JZ2oC82K19P5bjcwWzEbuHuwjUU5RN0SZarOU6xWc3C7+TWzqr6CR4ZgF3qwX4AK6rqfsFqrOkN9/8=
X-Received: by 2002:a1c:230f:: with SMTP id j15mr4909808wmj.100.1591396013500;  Fri, 05 Jun 2020 15:26:53 -0700 (PDT)
MIME-Version: 1.0
References: <CALGR9oZb7aTyFSFr-wg8-v5=rBgGPE0=RMj6Xed_HwPVoyoEOw@mail.gmail.com>
In-Reply-To: <CALGR9oZb7aTyFSFr-wg8-v5=rBgGPE0=RMj6Xed_HwPVoyoEOw@mail.gmail.com>
From: Lucas Pardue <lucaspardue.24.7@gmail.com>
Date: Fri, 5 Jun 2020 23:26:42 +0100
Message-ID: <CALGR9oYHLcTyhZR_3=T6HXq3QqkWpxoK2oOA407jLXTmz=DUtQ@mail.gmail.com>
Subject: Re: Consensus call for Late-stage documents, June 2020 edition
To: QUIC WG <quic@ietf.org>
Cc: Lars Eggert <lars@eggert.org>, Mark Nottingham <mnot@mnot.net>
Content-Type: multipart/alternative; boundary="00000000000014ee3505a75dc273"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/5KCBzSk_6R3eWWJE5AVliwZpR90>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 05 Jun 2020 22:26:57 -0000

--00000000000014ee3505a75dc273
Content-Type: text/plain; charset="UTF-8"

Hearing no pushback, consensus is declared.

Regards
Lars, Lucas and Mark
QUIC WG Chairs

--00000000000014ee3505a75dc273
Content-Type: text/html; charset="UTF-8"

<div dir="ltr"><div>Hearing no pushback, consensus is declared.</div><div><br></div><div>Regards</div><div>
Lars, Lucas and Mark<br> QUIC WG Chairs

</div></div>

--00000000000014ee3505a75dc273--


From nobody Sun Jun  7 00:32:38 2020
Return-Path: <do_not_reply@mnot.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id DF5B93A09FB for <quic@ietfa.amsl.com>; Sun,  7 Jun 2020 00:32:36 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, HTML_MESSAGE=0.001, RCVD_IN_MSPIKE_H3=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=mnot.net header.b=cKIUMujm; dkim=pass (2048-bit key) header.d=messagingengine.com header.b=ncYzAKKt
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id kqP2sSooxJbU for <quic@ietfa.amsl.com>; Sun,  7 Jun 2020 00:32:32 -0700 (PDT)
Received: from out1-smtp.messagingengine.com (out1-smtp.messagingengine.com [66.111.4.25]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 6F7B83A09FA for <quic@ietf.org>; Sun,  7 Jun 2020 00:32:32 -0700 (PDT)
Received: from compute1.internal (compute1.nyi.internal [10.202.2.41]) by mailout.nyi.internal (Postfix) with ESMTP id 9AE535C0096 for <quic@ietf.org>; Sun,  7 Jun 2020 03:32:30 -0400 (EDT)
Received: from mailfrontend2 ([10.202.2.163]) by compute1.internal (MEProxy); Sun, 07 Jun 2020 03:32:30 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=mnot.net; h= content-type:mime-version:from:to:subject:message-id:date; s= fm2; bh=NP68HsRUkCz20eHUKyUWVYC5tRqUcLpDUVoTq7HkoMw=; b=cKIUMujm oC6q+FCA+K1K8jei1dpKcIFN4H+ZAFCeOl8JVZVHy6FoOFmX1UtkncYjRbKuiaB6 v1D2Oi4V+7xcSN3+BmPu/m/Kv5jfNZ9K2VjWEMulMHXxEGzVtvAXn5cHME60SUaf NICyi5l81jwRkhwrEb9X+Hu/iWNCJFGS+Rro1tF83D3iD9S9UIefn6lqBI4bxzyo PxpgeLmv0DlSHZI6jgFgCSxGxu2c1V2P1f023w+eGhWOfKJl32jC9v5BnWvwUow5 sMKhovbuLqdavdeXmpvOf8i7lRR86CWSNjqfH0QMiMgbmVkNLIBrIBhkEg+p688+ /Qw4fPI46yODtg==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d= messagingengine.com; h=content-type:date:from:message-id :mime-version:subject:to:x-me-proxy:x-me-proxy:x-me-sender :x-me-sender:x-sasl-enc; s=fm3; bh=NP68HsRUkCz20eHUKyUWVYC5tRqUc LpDUVoTq7HkoMw=; b=ncYzAKKtPwfSilA1GWCdZiaURLklToJrvlXXJZEi2pAil SoZOxVU7EQSTPLXlvMDRrQHwKfwJdAR8ztrqO4NDcyuWGg1IZjJ2HmamZZE8csMy P3eVyVCV94oH9j7yqRlJ1Vmxbo0lH8Oi6JDtYN3yrJQmNZwpyAJUO4kA7H/9/2ED r6BKoqFSk7WXgdbQxSLloDPdhA2XIAA97luMCIbqRvlU68sL9USIMl0ZAV8dmqrh 28FOXIr5fSqIka29lMUegpuXNatleSHQCCBmdOj2ZiRMJ3JKtmqhjNt8CrkpCdU8 k/GuE7vTTQjb6XkCbxHu3sRBh3rrfJq2rqM5tnS9Q==
X-ME-Sender: <xms:DpjcXp7kZgxnxEuZ-qV06LjKeL-Yl5mrLOoP9ijAPxIKjBIUy5cBHg>
X-ME-Proxy-Cause: gggruggvucftvghtrhhoucdtuddrgeduhedrudegkedgvdeiucetufdoteggodetrfdotf fvucfrrhhofhhilhgvmecuhfgrshhtofgrihhlpdfqfgfvpdfurfetoffkrfgpnffqhgen uceurghilhhouhhtmecufedttdenucenucfjughrpegtggfhvffusegrtddtredttdejne cuhfhrohhmpeftvghpohhsihhtohhrhicutegtthhivhhithihucfuuhhmmhgrrhihuceu ohhtuceoughopghnohhtpghrvghplhihsehmnhhothdrnhgvtheqnecuggftrfgrthhtvg hrnhepkeefvdduteejvdefkeehieevuefgfefhteetveegffekffefteffvdelheduieet necuffhomhgrihhnpehgihhthhhusgdrtghomhenucfkphepgedtrdeihedruddufedrud efvdenucevlhhushhtvghrufhiiigvpedtnecurfgrrhgrmhepmhgrihhlfhhrohhmpegu ohgpnhhothgprhgvphhlhiesmhhnohhtrdhnvght
X-ME-Proxy: <xmx:DpjcXm5n08bTx0HGXQE_jEYg2Vq830F0dZE8zmmtzrpKvTfo9HLZpA> <xmx:DpjcXgc2uEvF5UtLuArJx1OLbf5QeXhXD8EYDwr9Z8dRzGU9-VGMBQ> <xmx:DpjcXiK3UOMYBOTlC2k6V_yDhYa0v3wIf_w1aA6aeEV8oYB1Rikjog> <xmx:DpjcXvxsnXxKCb0nDwQ3OUCkP7i7AF8364tybF0zZp3GQ0B1NRYIew>
Received: from fv-az115.internal.cloudapp.net (unknown [40.65.113.132]) by mail.messagingengine.com (Postfix) with ESMTPA id 0EC3E3060F09 for <quic@ietf.org>; Sun,  7 Jun 2020 03:32:29 -0400 (EDT)
Content-Type: multipart/alternative; boundary="===============0272525925448415357=="
MIME-Version: 1.0
From: Repository Activity Summary Bot <do_not_reply@mnot.net>
To: quic@ietf.org
Subject: Weekly github digest (QUIC Activity Summary)
Message-Id: <20200607073230.0EC3E3060F09@mailuser.nyi.internal>
Date: Sun,  7 Jun 2020 03:32:29 -0400 (EDT)
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/3aSnD1aH2ocQn61eQUQ8WLjtZCo>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Sun, 07 Jun 2020 07:32:37 -0000

--===============0272525925448415357==
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain; charset="utf-8"; format="flowed"




Events without label "editorial"

Issues
------
* quicwg/base-drafts (+18/-0/=F0=9F=92=AC35)
  18 issues created:
  - max_udp_payload_size (0x03) needs a ref. to packet size section. (by go=
rryfair)
    https://github.com/quicwg/base-drafts/issues/3738=20
  - Frame Types and Formats by transport requirements (by gorryfair)
    https://github.com/quicwg/base-drafts/issues/3737=20
  - bundling and coalescing? (by gorryfair)
    https://github.com/quicwg/base-drafts/issues/3736=20
  - QUIC transport Reading NiTs part 2 of 2 (by gorryfair)
    https://github.com/quicwg/base-drafts/issues/3735=20
  - Three ECN NiTS (by gorryfair)
    https://github.com/quicwg/base-drafts/issues/3734=20
  - Question on ECN validation (by larseggert)
    https://github.com/quicwg/base-drafts/issues/3733 [-transport]=20
  - Gorrys invariant nits (by gorryfair)
    https://github.com/quicwg/base-drafts/issues/3731=20
  - RFC2119 keyword appears within a note (by gorryfair)
    https://github.com/quicwg/base-drafts/issues/3729=20
  - Packet numbers and sequence numbers are too easily confused (by jonglez=
b)
    https://github.com/quicwg/base-drafts/issues/3725=20
  - Different meanings of "message" are mixed up in transport draft (by jon=
glezb)
    https://github.com/quicwg/base-drafts/issues/3724=20
  - HTTP: provide guidance on PING usage (by dtikhonov)
    https://github.com/quicwg/base-drafts/issues/3723=20
  - QUIC Transport: Flow Control (by gorryfair)
    https://github.com/quicwg/base-drafts/issues/3722=20
  - QUIC Transport - sending after final size? (by gorryfair)
    https://github.com/quicwg/base-drafts/issues/3721=20
  - QUIC Transport: Reading NiTs - set 1 (by gorryfair)
    https://github.com/quicwg/base-drafts/issues/3720=20
  - Tweak the split between quic-transport and quic-tls drafts (by DavidSch=
inazi)
    https://github.com/quicwg/base-drafts/issues/3717=20
  - MUST not -> MUST NOT (by larseggert)
    https://github.com/quicwg/base-drafts/issues/3715=20
  - Are both types of CONNECTION_CLOSE frames permitted during the handshak=
e? (by rpaulo)
    https://github.com/quicwg/base-drafts/issues/3713=20
  - Make SERVER_BUSY more generic (by martinthomson)
    https://github.com/quicwg/base-drafts/issues/3709 [-transport]=20

  13 issues received 35 new comments:
  - #3736 bundling and coalescing? (1 by MikeBishop)
    https://github.com/quicwg/base-drafts/issues/3736 [-transport]=20
  - #3733 Question on ECN validation (6 by larseggert, mirjak)
    https://github.com/quicwg/base-drafts/issues/3733 [-transport]=20
  - #3725 Packet numbers and sequence numbers are too easily confused (1 by=
 ianswett)
    https://github.com/quicwg/base-drafts/issues/3725=20
  - #3724 Different meanings of "message" are mixed up in transport draft (=
1 by martinthomson)
    https://github.com/quicwg/base-drafts/issues/3724=20
  - #3723 HTTP: provide guidance on PING usage (2 by MikeBishop, mirjak)
    https://github.com/quicwg/base-drafts/issues/3723 [-http]=20
  - #3722 QUIC Transport: Flow Control (1 by evolodina)
    https://github.com/quicwg/base-drafts/issues/3722 [-transport]=20
  - #3721 QUIC Transport - sending after final size? (2 by MikeBishop, gorr=
yfair)
    https://github.com/quicwg/base-drafts/issues/3721 [-transport]=20
  - #3717 Tweak the split between quic-transport and quic-tls drafts (2 by =
chris-wood, ianswett)
    https://github.com/quicwg/base-drafts/issues/3717 [-tls] [-transport]=20
  - #3713 Are both types of CONNECTION_CLOSE frames permitted during the ha=
ndshake? (4 by janaiyengar, kazuho, martinthomson, rpaulo)
    https://github.com/quicwg/base-drafts/issues/3713=20
  - #3709 Make SERVER_BUSY more generic (5 by MikeBishop, janaiyengar, kazu=
ho, martinthomson, mjoras)
    https://github.com/quicwg/base-drafts/issues/3709 [-transport] [design]=
 [proposal-ready]=20
  - #3690 How to reject a connection attempt (1 by mjoras)
    https://github.com/quicwg/base-drafts/issues/3690 [-transport]=20
  - #3661 Include epoch in the AAD or the nonce? (8 by DavidSchinazi, ekr, =
fxguenther, ianswett, kazuho, larseggert, martinthomson)
    https://github.com/quicwg/base-drafts/issues/3661 [-tls] [design] [prop=
osal-ready]=20
  - #3529 Changing the Default QUIC ACK Policy (1 by LPardue)
    https://github.com/quicwg/base-drafts/issues/3529 [-recovery] [-transpo=
rt] [design]=20

* quicwg/ops-drafts (+1/-0/=F0=9F=92=AC1)
  1 issues created:
  - Indicate which parts of the draft are v1 specific (by mirjak)
    https://github.com/quicwg/ops-drafts/issues/95=20

  1 issues received 1 new comments:
  - #75 Document how SNI from QUIC handshakes is presented (1 by mirjak)
    https://github.com/quicwg/ops-drafts/issues/75 [-manageability]=20



Pull requests
-------------
* quicwg/base-drafts (+5/-1/=F0=9F=92=AC10)
  5 pull requests submitted:
  - Minor editorializing of CID authentication (by janaiyengar)
    https://github.com/quicwg/base-drafts/pull/3732=20
  - Gorrys invariant nits (by gorryfair)
    https://github.com/quicwg/base-drafts/pull/3730=20
  - Add missing Length field to header protection example. (by kixelated)
    https://github.com/quicwg/base-drafts/pull/3728=20
  - Fix typo (by janaiyengar)
    https://github.com/quicwg/base-drafts/pull/3716=20
  - Minor editorial fix (by janaiyengar)
    https://github.com/quicwg/base-drafts/pull/3710=20

  3 pull requests received 10 new comments:
  - #3730 Gorrys invariant nits (1 by larseggert)
    https://github.com/quicwg/base-drafts/pull/3730=20
  - #3710 Minor editorial fix (7 by MikeBishop, janaiyengar, larseggert, ma=
rtinthomson)
    https://github.com/quicwg/base-drafts/pull/3710=20
  - #3702 DPLPMTU merge tweaks (2 by gorryfair)
    https://github.com/quicwg/base-drafts/pull/3702 [-transport] [design]=20

  1 pull requests merged:
  - Fix typo
    https://github.com/quicwg/base-drafts/pull/3716=20

* quicwg/ops-drafts (+0/-0/=F0=9F=92=AC1)
  1 pull requests received 1 new comments:
  - #94 DDoS mitigations - linking after NAT rebinding and UDP Policing (1 =
by igorlord)
    https://github.com/quicwg/ops-drafts/pull/94=20


Repositories tracked by this digest:
-----------------------------------
* https://github.com/quicwg/base-drafts
* https://github.com/quicwg/ops-drafts
* https://github.com/quicwg/datagram
* https://github.com/quicwg/load-balancers
* https://github.com/quicwg/version-negotiation

--===============0272525925448415357==
Content-Type: text/html; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable

<!doctype html>
<html lang=3D"en">
<head>
<meta charset=3D"utf-8">
<title>Weekly github digest (QUIC Activity Summary)</title>
<style>
body { font-family: Gotham, "Helvetica Neue", Helvetica, Arial, sans-serif;=
 font-size: 14px; }
h2 { margin-top: 3em; color: #A52A2A; font-style: italic; font-weight: norm=
al; }
h3 { margin-bottom:0; margin-top: 2em; font-size: 1.2em; }
h1+h2 { margin-top: 1em; }
a { color: #bb6219; text-decoration: none; }
li { margin-bottom: .35em; }
.repos { margin-bottom: 0; margin-top:0; line-height: 1.2; }
.new { color: red; }
.label { display: inline;
	padding: .2em .6em .3em;
	font-size: 75%;
	font-weight: 700;
	line-height: 1;
	color: #fff;
	text-align: center;
	white-space: nowrap;
	vertical-align: baseline;
	border-radius: .25em;
}
</style>
</head>

<body>
<h1>Sunday June 07, 2020</h1>

<p>Events without label "editorial"</p>

<h2>Issues</h2>

<h3>quicwg/base-drafts (+18/-0/=F0=9F=92=AC35)</h3>
  <p class=3D"new">18 issues created:</p>
  <ul>
  <li>#3738 <a href=3D"https://github.com/quicwg/base-drafts/issues/3738">m=
ax_udp_payload_size (0x03) needs a ref. to packet size section.</a> (by gor=
ryfair) </li>
 =20
  <li>#3737 <a href=3D"https://github.com/quicwg/base-drafts/issues/3737">F=
rame Types and Formats by transport requirements</a> (by gorryfair) </li>
 =20
  <li>#3736 <a href=3D"https://github.com/quicwg/base-drafts/issues/3736">b=
undling and coalescing?</a> (by gorryfair) </li>
 =20
  <li>#3735 <a href=3D"https://github.com/quicwg/base-drafts/issues/3735">Q=
UIC transport Reading NiTs part 2 of 2</a> (by gorryfair) </li>
 =20
  <li>#3734 <a href=3D"https://github.com/quicwg/base-drafts/issues/3734">T=
hree ECN NiTS</a> (by gorryfair) </li>
 =20
  <li>#3733 <a href=3D"https://github.com/quicwg/base-drafts/issues/3733">Q=
uestion on ECN validation</a> (by larseggert) <span class=3D"label" style=
=3D"background-color: #fbca04; color: #000000">-transport</span> </li>
 =20
  <li>#3731 <a href=3D"https://github.com/quicwg/base-drafts/issues/3731">G=
orrys invariant nits</a> (by gorryfair) </li>
 =20
  <li>#3729 <a href=3D"https://github.com/quicwg/base-drafts/issues/3729">R=
FC2119 keyword appears within a note</a> (by gorryfair) </li>
 =20
  <li>#3725 <a href=3D"https://github.com/quicwg/base-drafts/issues/3725">P=
acket numbers and sequence numbers are too easily confused</a> (by jonglezb=
) </li>
 =20
  <li>#3724 <a href=3D"https://github.com/quicwg/base-drafts/issues/3724">D=
ifferent meanings of &quot;message&quot; are mixed up in transport draft</a=
> (by jonglezb) </li>
 =20
  <li>#3723 <a href=3D"https://github.com/quicwg/base-drafts/issues/3723">H=
TTP: provide guidance on PING usage</a> (by dtikhonov) </li>
 =20
  <li>#3722 <a href=3D"https://github.com/quicwg/base-drafts/issues/3722">Q=
UIC Transport: Flow Control</a> (by gorryfair) </li>
 =20
  <li>#3721 <a href=3D"https://github.com/quicwg/base-drafts/issues/3721">Q=
UIC Transport - sending after final size?</a> (by gorryfair) </li>
 =20
  <li>#3720 <a href=3D"https://github.com/quicwg/base-drafts/issues/3720">Q=
UIC Transport: Reading NiTs - set 1</a> (by gorryfair) </li>
 =20
  <li>#3717 <a href=3D"https://github.com/quicwg/base-drafts/issues/3717">T=
weak the split between quic-transport and quic-tls drafts</a> (by DavidSchi=
nazi) </li>
 =20
  <li>#3715 <a href=3D"https://github.com/quicwg/base-drafts/issues/3715">M=
UST not -&gt; MUST NOT</a> (by larseggert) </li>
 =20
  <li>#3713 <a href=3D"https://github.com/quicwg/base-drafts/issues/3713">A=
re both types of CONNECTION_CLOSE frames permitted during the handshake?</a=
> (by rpaulo) </li>
 =20
  <li>#3709 <a href=3D"https://github.com/quicwg/base-drafts/issues/3709">M=
ake SERVER_BUSY more generic</a> (by martinthomson) <span class=3D"label" s=
tyle=3D"background-color: #fbca04; color: #000000">-transport</span> </li>
  </ul>

  <p>13 issues received 35 new comments:</p>
  <ul>
  <li>#3736 <a href=3D"https://github.com/quicwg/base-drafts/issues/3736">b=
undling and coalescing?</a> (1 by MikeBishop) <span class=3D"label" style=
=3D"background-color: #fbca04; color: #000000">-transport</span> </li>
 =20
  <li>#3733 <a href=3D"https://github.com/quicwg/base-drafts/issues/3733">Q=
uestion on ECN validation</a> (6 by larseggert, mirjak) <span class=3D"labe=
l" style=3D"background-color: #fbca04; color: #000000">-transport</span> </=
li>
 =20
  <li>#3725 <a href=3D"https://github.com/quicwg/base-drafts/issues/3725">P=
acket numbers and sequence numbers are too easily confused</a> (1 by ianswe=
tt) </li>
 =20
  <li>#3724 <a href=3D"https://github.com/quicwg/base-drafts/issues/3724">D=
ifferent meanings of &quot;message&quot; are mixed up in transport draft</a=
> (1 by martinthomson) </li>
 =20
  <li>#3723 <a href=3D"https://github.com/quicwg/base-drafts/issues/3723">H=
TTP: provide guidance on PING usage</a> (2 by MikeBishop, mirjak) <span cla=
ss=3D"label" style=3D"background-color: #fbca04; color: #000000">-http</spa=
n> </li>
 =20
  <li>#3722 <a href=3D"https://github.com/quicwg/base-drafts/issues/3722">Q=
UIC Transport: Flow Control</a> (1 by evolodina) <span class=3D"label" styl=
e=3D"background-color: #fbca04; color: #000000">-transport</span> </li>
 =20
  <li>#3721 <a href=3D"https://github.com/quicwg/base-drafts/issues/3721">Q=
UIC Transport - sending after final size?</a> (2 by MikeBishop, gorryfair) =
<span class=3D"label" style=3D"background-color: #fbca04; color: #000000">-=
transport</span> </li>
 =20
  <li>#3717 <a href=3D"https://github.com/quicwg/base-drafts/issues/3717">T=
weak the split between quic-transport and quic-tls drafts</a> (2 by chris-w=
ood, ianswett) <span class=3D"label" style=3D"background-color: #fbca04; co=
lor: #000000">-tls</span> <span class=3D"label" style=3D"background-color: =
#fbca04; color: #000000">-transport</span> </li>
 =20
  <li>#3713 <a href=3D"https://github.com/quicwg/base-drafts/issues/3713">A=
re both types of CONNECTION_CLOSE frames permitted during the handshake?</a=
> (4 by janaiyengar, kazuho, martinthomson, rpaulo) </li>
 =20
  <li>#3709 <a href=3D"https://github.com/quicwg/base-drafts/issues/3709">M=
ake SERVER_BUSY more generic</a> (5 by MikeBishop, janaiyengar, kazuho, mar=
tinthomson, mjoras) <span class=3D"label" style=3D"background-color: #fbca0=
4; color: #000000">-transport</span> <span class=3D"label" style=3D"backgro=
und-color: #1d76db; color: #ffffff">design</span> <span class=3D"label" sty=
le=3D"background-color: #c5def5; color: #000000">proposal-ready</span> </li>
 =20
  <li>#3690 <a href=3D"https://github.com/quicwg/base-drafts/issues/3690">H=
ow to reject a connection attempt</a> (1 by mjoras) <span class=3D"label" s=
tyle=3D"background-color: #fbca04; color: #000000">-transport</span> </li>
 =20
  <li>#3661 <a href=3D"https://github.com/quicwg/base-drafts/issues/3661">I=
nclude epoch in the AAD or the nonce?</a> (8 by DavidSchinazi, ekr, fxguent=
her, ianswett, kazuho, larseggert, martinthomson) <span class=3D"label" sty=
le=3D"background-color: #fbca04; color: #000000">-tls</span> <span class=3D=
"label" style=3D"background-color: #1d76db; color: #ffffff">design</span> <=
span class=3D"label" style=3D"background-color: #c5def5; color: #000000">pr=
oposal-ready</span> </li>
 =20
  <li>#3529 <a href=3D"https://github.com/quicwg/base-drafts/issues/3529">C=
hanging the Default QUIC ACK Policy</a> (1 by LPardue) <span class=3D"label=
" style=3D"background-color: #fbca04; color: #000000">-recovery</span> <spa=
n class=3D"label" style=3D"background-color: #fbca04; color: #000000">-tran=
sport</span> <span class=3D"label" style=3D"background-color: #1d76db; colo=
r: #ffffff">design</span> </li>
  </ul>


<h3>quicwg/ops-drafts (+1/-0/=F0=9F=92=AC1)</h3>
  <p class=3D"new">1 issues created:</p>
  <ul>
  <li>#95 <a href=3D"https://github.com/quicwg/ops-drafts/issues/95">Indica=
te which parts of the draft are v1 specific</a> (by mirjak) </li>
  </ul>

  <p>1 issues received 1 new comments:</p>
  <ul>
  <li>#75 <a href=3D"https://github.com/quicwg/ops-drafts/issues/75">Docume=
nt how SNI from QUIC handshakes is presented</a> (1 by mirjak) <span class=
=3D"label" style=3D"background-color: #fbca04; color: #000000">-manageabili=
ty</span> </li>
  </ul>




<h2>Pull requests</h2>
<h3>quicwg/base-drafts (+5/-1/=F0=9F=92=AC10)</h3>
  <p class=3D"new">5 pull requests submitted:</p>
  <ul>
  <li>#3732 <a href=3D"https://github.com/quicwg/base-drafts/pull/3732">Min=
or editorializing of CID authentication</a> (by janaiyengar) </li>
 =20
  <li>#3730 <a href=3D"https://github.com/quicwg/base-drafts/pull/3730">Gor=
rys invariant nits</a> (by gorryfair) </li>
 =20
  <li>#3728 <a href=3D"https://github.com/quicwg/base-drafts/pull/3728">Add=
 missing Length field to header protection example.</a> (by kixelated) </li>
 =20
  <li>#3716 <a href=3D"https://github.com/quicwg/base-drafts/pull/3716">Fix=
 typo</a> (by janaiyengar) </li>
 =20
  <li>#3710 <a href=3D"https://github.com/quicwg/base-drafts/pull/3710">Min=
or editorial fix</a> (by janaiyengar) </li>
  </ul>

  <p>3 pull requests received 10 new comments:</p>
  <ul>
  <li>#3730 <a href=3D"https://github.com/quicwg/base-drafts/pull/3730">Gor=
rys invariant nits</a> (1 by larseggert) </li>
 =20
  <li>#3710 <a href=3D"https://github.com/quicwg/base-drafts/pull/3710">Min=
or editorial fix</a> (7 by MikeBishop, janaiyengar, larseggert, martinthoms=
on) </li>
 =20
  <li>#3702 <a href=3D"https://github.com/quicwg/base-drafts/pull/3702">DPL=
PMTU merge tweaks</a> (2 by gorryfair) <span class=3D"label" style=3D"backg=
round-color: #fbca04; color: #000000">-transport</span> <span class=3D"labe=
l" style=3D"background-color: #1d76db; color: #ffffff">design</span> </li>
  </ul>

  <p>1 pull requests merged:</p>
  <ul>
  <li>#3716 <a href=3D"https://github.com/quicwg/base-drafts/pull/3716">Fix=
 typo</a> </li>
  </ul>

<h3>quicwg/ops-drafts (+0/-0/=F0=9F=92=AC1)</h3>

  <p>1 pull requests received 1 new comments:</p>
  <ul>
  <li>#94 <a href=3D"https://github.com/quicwg/ops-drafts/pull/94">DDoS mit=
igations - linking after NAT rebinding and UDP Policing</a> (1 by igorlord)=
 </li>
  </ul>



<h2>Repositories tracked by this digest:</h2>
<ul class=3D"repos">
  <li><a href=3D"https://github.com/quicwg/base-drafts">https://github.com/=
quicwg/base-drafts</a></li>
  <li><a href=3D"https://github.com/quicwg/ops-drafts">https://github.com/q=
uicwg/ops-drafts</a></li>
  <li><a href=3D"https://github.com/quicwg/datagram">https://github.com/qui=
cwg/datagram</a></li>
  <li><a href=3D"https://github.com/quicwg/load-balancers">https://github.c=
om/quicwg/load-balancers</a></li>
  <li><a href=3D"https://github.com/quicwg/version-negotiation">https://git=
hub.com/quicwg/version-negotiation</a></li>
  </ul>
</body>
</html>

--===============0272525925448415357==--


From nobody Mon Jun  8 14:01:20 2020
Return-Path: <martin.h.duke@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 0BC9F3A102C for <quic@ietfa.amsl.com>; Mon,  8 Jun 2020 14:01:08 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id H6tw97g3Me7S for <quic@ietfa.amsl.com>; Mon,  8 Jun 2020 14:01:06 -0700 (PDT)
Received: from mail-io1-xd29.google.com (mail-io1-xd29.google.com [IPv6:2607:f8b0:4864:20::d29]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id E7A443A1029 for <quic@ietf.org>; Mon,  8 Jun 2020 14:01:03 -0700 (PDT)
Received: by mail-io1-xd29.google.com with SMTP id r77so7390935ior.3 for <quic@ietf.org>; Mon, 08 Jun 2020 14:01:03 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:from:date:message-id:subject:to; bh=5nWu3TguOUxDiAx1SyWc5OLICDnngAIfFJnJG9iJMgQ=; b=HNBMXhYwRMU2DKDAmgqRPjSfDXB0AJDY1ceaF1YUSJ9fQmmPjO3VPssBbx6LJcalhG MbBq2ByfBZ1dSFzgd0lUUKV7c6YM1B9wjw5O62nQNlRj5ZlLwTNqkl4z5MlR/x89xkmh svl0XwxzpoY+2pRAc3dwvwz6H94SgO4dK1ixqVbSxYwoEg9tvPUKn8BN72QqDu5T1+EP jUQHKvfFOk+ZYhy7LrYMLldhqnn9XQqiZKMHNd9wJnYc0xYQXB8OcWc6hVstmVtM4EvA XwsqaJKxubAbjtrGZnNDKAjt2N6ECsddwFqiYVB9GPSnURrreB8qI10GiIXhDYYpJju4 LmwA==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:from:date:message-id:subject:to; bh=5nWu3TguOUxDiAx1SyWc5OLICDnngAIfFJnJG9iJMgQ=; b=nvhSrHfkCE7gsfv/vK7BnRufj6iTOSYxB8CFM0P9p0rq+t5xoA4OXETYFhNwvckiBF NdS64Ykg11r0S9MRNAdMtdzlr0m12KsnPVgJhlek1cpVtSZPRylpU4VDLCGbHyezVn4U lgNNQtaLv3i4jhiOXlKKvsNQFYdJ+Pc+qoi0fMr1Lc7aKo92XAV1OTXgUBxAlJ6ma0Xi 3/c1J+6oSyxRJjwYkRXHW659qIjQ2BaaM9K3ya6DKgMrNofBixH6F0faV7KsahZ7jWO+ V+GuSTdq/8fPWz2Z8o4aB252yBcmZgdGVwp8cCwCJn3vW/uyipLpgFJunffxxD0sYbe2 m21A==
X-Gm-Message-State: AOAM531cre0Z4Jvs30cU3MtOGOeNyY/yeLQ7kNN3M42Jpg44aMm8SuxN pgtoalEVrxIJb0EkPSX24MaOwR/QMOQUcaxLou6UKcpHi/8=
X-Google-Smtp-Source: ABdhPJywvDRLLtQU1KcfXm8ctqie/vgAWgcrWyx7GRLGMDKeeZjQQTKSrquWM/QKt/Lglpdx+FGR2LljiMDh1bGBtjY=
X-Received: by 2002:a02:dc8:: with SMTP id 191mr23900350jax.95.1591650063006;  Mon, 08 Jun 2020 14:01:03 -0700 (PDT)
MIME-Version: 1.0
From: Martin Duke <martin.h.duke@gmail.com>
Date: Mon, 8 Jun 2020 14:00:52 -0700
Message-ID: <CAM4esxSg=oWNLUoNK6Rg6x1KL5NvtWz8qyvk3aa-UWL3HtaodA@mail.gmail.com>
Subject: Conflicting requirements?
To: IETF QUIC WG <quic@ietf.org>
Content-Type: multipart/alternative; boundary="0000000000009cc5c805a798e893"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/A3EHHYIMixurQRTSjkzZzeQHjPE>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 08 Jun 2020 21:01:15 -0000

--0000000000009cc5c805a798e893
Content-Type: text/plain; charset="UTF-8"

In section 8.1 of quic-transport,
When using ALPN, endpoints MUST immediately close a connection (see Section
10.3 of [QUIC-TRANSPORT
<https://quicwg.org/base-drafts/draft-ietf-quic-tls.html#QUIC-TRANSPORT>])
with a no_application_protocol TLS alert (QUIC error code 0x178; see Section
4.10 <https://quicwg.org/base-drafts/draft-ietf-quic-tls.html#tls-errors>)
if an application protocol is not negotiated.

In 4.10 of quic-tls,
Endpoints MAY use a generic error code to avoid possibly exposing
confidential information.

Which one takes precedence? May the server send a generic error when there
is no ALPN code?

--0000000000009cc5c805a798e893
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">In section 8.1 of quic-transport,=C2=A0<div><span style=3D=
"font-family:Lora,serif;font-size:16px">When using ALPN, endpoints MUST imm=
ediately close a connection (see Section 10.3 of=C2=A0</span><span style=3D=
"font-family:Lora,serif;font-size:16px">[<a href=3D"https://quicwg.org/base=
-drafts/draft-ietf-quic-tls.html#QUIC-TRANSPORT" class=3D"gmail-xref" style=
=3D"text-decoration-line:none;white-space:nowrap">QUIC-TRANSPORT</a>]</span=
><span style=3D"font-family:Lora,serif;font-size:16px">) with a no_applicat=
ion_protocol TLS alert (QUIC error code 0x178; see=C2=A0</span><a href=3D"h=
ttps://quicwg.org/base-drafts/draft-ietf-quic-tls.html#tls-errors" class=3D=
"gmail-xref" style=3D"text-decoration-line:none;white-space:nowrap;font-fam=
ily:Lora,serif;font-size:16px">Section 4.10</a><span style=3D"font-family:L=
ora,serif;font-size:16px">) if an application protocol is not negotiated.</=
span>=C2=A0</div><div><br></div><div>In 4.10 of quic-tls,</div><div><span s=
tyle=3D"font-family:Lora,serif;font-size:16px">Endpoints MAY use a generic =
error code to avoid possibly exposing confidential information.</span></div=
><div><br></div><div>Which one takes precedence? May the server send a gene=
ric error when there is no ALPN code?=C2=A0<br></div></div>

--0000000000009cc5c805a798e893--


From nobody Mon Jun  8 14:13:09 2020
Return-Path: <huitema@huitema.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 82A233A0063 for <quic@ietfa.amsl.com>; Mon,  8 Jun 2020 14:13:06 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.898
X-Spam-Level: 
X-Spam-Status: No, score=-1.898 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id h7jkzkEoeEIg for <quic@ietfa.amsl.com>; Mon,  8 Jun 2020 14:12:59 -0700 (PDT)
Received: from mx36-out10.antispamcloud.com (mx36-out10.antispamcloud.com [209.126.121.30]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 62D6F3A0062 for <quic@ietf.org>; Mon,  8 Jun 2020 14:12:59 -0700 (PDT)
Received: from xse174.mail2web.com ([66.113.196.174] helo=xse.mail2web.com) by mx168.antispamcloud.com with esmtp (Exim 4.92) (envelope-from <huitema@huitema.net>) id 1jiP4c-000hOA-F1 for quic@ietf.org; Mon, 08 Jun 2020 23:12:58 +0200
Received: from xsmtp22.mail2web.com (unknown [10.100.68.61]) by xse.mail2web.com (Postfix) with ESMTPS id 49gmDW2d9dz2tm for <quic@ietf.org>; Mon,  8 Jun 2020 14:12:43 -0700 (PDT)
Received: from [10.5.2.12] (helo=xmail02.myhosting.com) by xsmtp22.mail2web.com with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:256) (Exim 4.92) (envelope-from <huitema@huitema.net>) id 1jiP4V-0003yD-7h for quic@ietf.org; Mon, 08 Jun 2020 14:12:43 -0700
Received: (qmail 12934 invoked from network); 8 Jun 2020 21:12:42 -0000
Received: from unknown (HELO [192.168.1.107]) (Authenticated-user:_huitema@huitema.net@[172.58.43.64]) (envelope-sender <huitema@huitema.net>) by xmail02.myhosting.com (qmail-ldap-1.03) with ESMTPA for <quic@ietf.org>; 8 Jun 2020 21:12:42 -0000
Subject: Re: Conflicting requirements?
To: Martin Duke <martin.h.duke@gmail.com>, IETF QUIC WG <quic@ietf.org>
References: <CAM4esxSg=oWNLUoNK6Rg6x1KL5NvtWz8qyvk3aa-UWL3HtaodA@mail.gmail.com>
From: Christian Huitema <huitema@huitema.net>
Autocrypt: addr=huitema@huitema.net; prefer-encrypt=mutual; keydata= mDMEXtavGxYJKwYBBAHaRw8BAQdA1ou9A5MHTP9N3jfsWzlDZ+jPnQkusmc7sfLmWVz1Rmu0 J0NocmlzdGlhbiBIdWl0ZW1hIDxodWl0ZW1hQGh1aXRlbWEubmV0PoiWBBMWCAA+FiEEw3G4 Nwi4QEpAAXUUELAmqKBYtJQFAl7WrxsCGwMFCQlmAYAFCwkIBwIGFQoJCAsCBBYCAwECHgEC F4AACgkQELAmqKBYtJQbMwD/ebj/qnSbthC/5kD5DxZ/Ip0CGJw5QBz/+fJp3R8iAlsBAMjK r2tmyWyJz0CUkVG24WaR5EAJDvgwDv8h22U6QVkAuDgEXtavGxIKKwYBBAGXVQEFAQEHQJoM 6MUAIqpoqdCIiACiEynZf7nlJg2Eu0pXIhbUGONdAwEIB4h+BBgWCAAmFiEEw3G4Nwi4QEpA AXUUELAmqKBYtJQFAl7WrxsCGwwFCQlmAYAACgkQELAmqKBYtJRm2wD7BzeK5gEXSmBcBf0j BYdSaJcXNzx4yPLbP4GnUMAyl2cBAJzcsR4RkwO4dCRqM9CHpVJCwHtbUDJaa55//E0kp+gH
Message-ID: <0cde7047-7856-97c5-5e3e-0f0ad380451a@huitema.net>
Date: Mon, 8 Jun 2020 14:12:41 -0700
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:68.0) Gecko/20100101 Thunderbird/68.8.0
MIME-Version: 1.0
In-Reply-To: <CAM4esxSg=oWNLUoNK6Rg6x1KL5NvtWz8qyvk3aa-UWL3HtaodA@mail.gmail.com>
Content-Type: multipart/alternative; boundary="------------595D21BEBB84B5F58E3E485E"
Content-Language: en-US
X-Originating-IP: 66.113.196.174
X-Spampanel-Domain: xsmtpout.mail2web.com
X-Spampanel-Username: 66.113.196.174/32
Authentication-Results: antispamcloud.com; auth=pass smtp.auth=66.113.196.174/32@xsmtpout.mail2web.com
X-Spampanel-Outgoing-Class: unsure
X-Spampanel-Outgoing-Evidence: Combined (0.15)
X-Recommended-Action: accept
X-Filter-ID: Mvzo4OR0dZXEDF/gcnlw0fwuhl+SFh+udCLiTeLN1rOpSDasLI4SayDByyq9LIhVUZbR67CQ7/vm /hHDJU4RXkTNWdUk1Ol2OGx3IfrIJKywOmJyM1qr8uRnWBrbSAGDcnqpk5VeF3xR4kF6iVwRtbgN zB/4Jkrw1eDLcif59ft0VWvQEKRy6njUpt9IaltjU7Tmz6iKnkQL9gqsxD347235Nhqq+/HvroPq 8GSPg+60/QPNqXybIny9WGhadIo/YxFBjiwvt8YGG8KnqhbQXbyme9ldZJ7uNXfg/GfS8fUvP/L5 rCqHDsKZM+xa1iwJX+gRCHfMVnsAk591zk0uilUI+ZL4xWiN8NS6C+dmX6OEdA4u1aThyWrQ/ou2 +v/lmX4Em37yFgrCB6NHRn1g+f3uncIqYSL3lhh5c81YyJqFoLZMmkWsaurVZfvqROaDnDtHb8z5 dpPkEuJ8Snwqla7jUnW3hy14Yji8fo+4xCnSRo4Rcu5Z37rMuDjCny5fE9ykbJ7I9co1MAEE3ruN Xsm8UJsAPvDcVSKtDCYkioPY5Qx4fJOk03R5fJtf/Dv/dkIzS7m4GUpXCY1Y3j3ilZpnfSx1WjM8 Y221ho/pD3/YTgm/NnC/c8YEQ5zoBxO0grdS7Rh5sH3gRcQbBXAoPEFgpch+Ue6WLYWFoyi0E6VU PVcmx1QL+XiKf76y/BgKM+kBC+3ISsPcm/xMvGjFNPKY2AXNZGS5G93aGyH8MqMlOQRMVMd0HCeT skOZ5TL8dyr00deay+U3cA4KLzNxeDXg724gFzhHYUe+7aKm0vVMjzgoJvC4aZvfZOucZxEfTi+J 2sBvM/O0p+zizleC4va6FPcpDHjXMKZJK8+chiaI0MTr3GW/OYY+JBcn69qp7cTs80/2FnZg/IMs IAdedSzLrjsyfTPCYbMCLdmf5h2vfxw3Qvb2Glio5Cia/9Kfg4kJ0WtAYbrpe3OOAtQNb87OBHCz Hbokiue7PjVB1S6AQRz4SqXhOP5fdiQt7lu5Jm5nk4BSgYHOJJgUtm67rBRli6kULE5BQDZnPvvF VsQ=
X-Report-Abuse-To: spam@quarantine11.antispamcloud.com
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/i_9vmzLYYXLlbJn6Mn3dNAgPJ9g>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 08 Jun 2020 21:13:07 -0000

This is a multi-part message in MIME format.
--------------595D21BEBB84B5F58E3E485E
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit

On 6/8/2020 2:00 PM, Martin Duke wrote:

> In section 8.1 of quic-transport, 
> When using ALPN, endpoints MUST immediately close a connection (see
> Section 10.3 of [QUIC-TRANSPORT
> <https://quicwg.org/base-drafts/draft-ietf-quic-tls.html#QUIC-TRANSPORT>])
> with a no_application_protocol TLS alert (QUIC error code 0x178;
> see Section 4.10
> <https://quicwg.org/base-drafts/draft-ietf-quic-tls.html#tls-errors>)
> if an application protocol is not negotiated. 
>
> In 4.10 of quic-tls,
> Endpoints MAY use a generic error code to avoid possibly exposing
> confidential information.
>
> Which one takes precedence? May the server send a generic error when
> there is no ALPN code?


That's a very old debate. More specific the error codes make debugging
easier but do expose details of implementation or configuration, so
there is a tension. The usual  compromise is that privacy conscious
applications can always be configured to send generic error codes.

-- Christian Huitema


--------------595D21BEBB84B5F58E3E485E
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: 8bit

<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <p>On 6/8/2020 2:00 PM, Martin Duke wrote:<br>
    </p>
    <blockquote type="cite"
cite="mid:CAM4esxSg=oWNLUoNK6Rg6x1KL5NvtWz8qyvk3aa-UWL3HtaodA@mail.gmail.com">
      <meta http-equiv="content-type" content="text/html; charset=UTF-8">
      <div dir="ltr">In section 8.1 of quic-transport, 
        <div><span style="font-family:Lora,serif;font-size:16px">When
            using ALPN, endpoints MUST immediately close a connection
            (see Section 10.3 of </span><span
            style="font-family:Lora,serif;font-size:16px">[<a
href="https://quicwg.org/base-drafts/draft-ietf-quic-tls.html#QUIC-TRANSPORT"
              class="gmail-xref"
              style="text-decoration-line:none;white-space:nowrap"
              moz-do-not-send="true">QUIC-TRANSPORT</a>]</span><span
            style="font-family:Lora,serif;font-size:16px">) with a
            no_application_protocol TLS alert (QUIC error code 0x178;
            see </span><a
href="https://quicwg.org/base-drafts/draft-ietf-quic-tls.html#tls-errors"
            class="gmail-xref"
style="text-decoration-line:none;white-space:nowrap;font-family:Lora,serif;font-size:16px"
            moz-do-not-send="true">Section 4.10</a><span
            style="font-family:Lora,serif;font-size:16px">) if an
            application protocol is not negotiated.</span> </div>
        <div><br>
        </div>
        <div>In 4.10 of quic-tls,</div>
        <div><span style="font-family:Lora,serif;font-size:16px">Endpoints
            MAY use a generic error code to avoid possibly exposing
            confidential information.</span></div>
        <div><br>
        </div>
        <div>Which one takes precedence? May the server send a generic
          error when there is no ALPN code? <br>
        </div>
      </div>
    </blockquote>
    <p><br>
    </p>
    <p>That's a very old debate. More specific the error codes make
      debugging easier but do expose details of implementation or
      configuration, so there is a tension. The usual  compromise is
      that privacy conscious applications can always be configured to
      send generic error codes.</p>
    <p>-- Christian Huitema<br>
    </p>
  </body>
</html>

--------------595D21BEBB84B5F58E3E485E--


From nobody Mon Jun  8 14:15:02 2020
Return-Path: <martin.h.duke@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id CE0763A00C4 for <quic@ietfa.amsl.com>; Mon,  8 Jun 2020 14:14:44 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 71W9XOlVTzQO for <quic@ietfa.amsl.com>; Mon,  8 Jun 2020 14:14:42 -0700 (PDT)
Received: from mail-io1-xd36.google.com (mail-io1-xd36.google.com [IPv6:2607:f8b0:4864:20::d36]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id D09653A00B2 for <quic@ietf.org>; Mon,  8 Jun 2020 14:14:42 -0700 (PDT)
Received: by mail-io1-xd36.google.com with SMTP id c8so20423115iob.6 for <quic@ietf.org>; Mon, 08 Jun 2020 14:14:42 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=KABJF9e84JET0N/tqqIYgIoZblHgUDpR9PXtFJMZ8Mg=; b=EtlNKiEL4Fnk/dwqc4VMzhKKR+HYYqBcNDbzT1H/h6BYFXXTwEvvELN93k+eC3T7uz G2aYg/3JeiTTNksh3lk0vPQXwXK64BVi8ZDfnf4qhy7A6xnx8Qhfh0xot4gXACXkwEA4 AaSE46GniIuzsMayQwPhq9Ott4tcAAf6O6KzRnRF9AhUJ1IJABNL9K98P/qW4cy4hw9Y SnMgVuS7O/LCH7qfz9k/Zc1tGCAfxbBr12gHdz51j7EAM66YQ4hSZPSBQ0gsUSqHtZPs ZoPzXk2IKgsq0+Nu1coMfy9KXozigsXLZwNevRrV7jhb+xaU2LSHEmeNi5KI6vL6kJkU B86g==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=KABJF9e84JET0N/tqqIYgIoZblHgUDpR9PXtFJMZ8Mg=; b=FrxVr1WXsM2tlbCA8Ci1+BRZI/ntBMhyw/F0QAzuFpF600dkJx2D/QiuN4qHlkwszR jMZ8yTULHKsPayPQN/TN9m6CiLOVrOufVuR3X2DwB5gfGmMxniz6wOrcE3NlJ/HIFcHS aRWWCJTrm2XGnMcJK+La9XGJNpQTWzzrDVQqF8JNdXnB94QwTyloN0uMj1k84damu9DY GSRMKRYjNz9jOymj92tFYJEP1bMK2FwruvFHqhfDTToSDJJ+Cj1lOzog6QjSccX8pdzI cWjU6d76VgtPQgYV1OwM+UyvW819cVGm4qn+0qpX1pEn4bnizJSDuxpO7cV9t642p8oq +U3w==
X-Gm-Message-State: AOAM530BjY4o1kmmG4auA9szRJ3rJw2arIwbD2WziamtbASJ2m6wiKQZ kF75rfAWaC4Z687WDhjsiXapFLCMKE+gqNEzIkzr1ufc
X-Google-Smtp-Source: ABdhPJzgxNItsoTiL1/HG3Trg6qp17+Ogl2Z2N3d9bF+16ZXlBqxyJ1IIPyK5Jadb5cjrm/WhcC1Vsy0mEARWn3/TkI=
X-Received: by 2002:a6b:fd18:: with SMTP id c24mr22368180ioi.97.1591650881651;  Mon, 08 Jun 2020 14:14:41 -0700 (PDT)
MIME-Version: 1.0
References: <CAM4esxSg=oWNLUoNK6Rg6x1KL5NvtWz8qyvk3aa-UWL3HtaodA@mail.gmail.com> <0cde7047-7856-97c5-5e3e-0f0ad380451a@huitema.net>
In-Reply-To: <0cde7047-7856-97c5-5e3e-0f0ad380451a@huitema.net>
From: Martin Duke <martin.h.duke@gmail.com>
Date: Mon, 8 Jun 2020 14:14:30 -0700
Message-ID: <CAM4esxQ+i=6d6U6iWTe5Ym=2F4-kt5BTqS-Bp=FPM1NBhnN+YA@mail.gmail.com>
Subject: Re: Conflicting requirements?
To: Christian Huitema <huitema@huitema.net>
Cc: IETF QUIC WG <quic@ietf.org>
Content-Type: multipart/alternative; boundary="00000000000068475005a7991941"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/y_YCC75R3k9X53vF2Zt7eMAwiSY>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 08 Jun 2020 21:14:45 -0000

--00000000000068475005a7991941
Content-Type: text/plain; charset="UTF-8"

I'm not debating the relative merits. We let our customers set this as they
wish. I'm just asking what the spec is requiring here, and if there's an
editorial change needed in one of the drafts.


On Mon, Jun 8, 2020 at 2:12 PM Christian Huitema <huitema@huitema.net>
wrote:

> On 6/8/2020 2:00 PM, Martin Duke wrote:
>
> In section 8.1 of quic-transport,
> When using ALPN, endpoints MUST immediately close a connection (see
> Section 10.3 of [QUIC-TRANSPORT
> <https://quicwg.org/base-drafts/draft-ietf-quic-tls.html#QUIC-TRANSPORT>])
> with a no_application_protocol TLS alert (QUIC error code 0x178; see Section
> 4.10 <https://quicwg.org/base-drafts/draft-ietf-quic-tls.html#tls-errors>)
> if an application protocol is not negotiated.
>
> In 4.10 of quic-tls,
> Endpoints MAY use a generic error code to avoid possibly exposing
> confidential information.
>
> Which one takes precedence? May the server send a generic error when there
> is no ALPN code?
>
>
> That's a very old debate. More specific the error codes make debugging
> easier but do expose details of implementation or configuration, so there
> is a tension. The usual  compromise is that privacy conscious applications
> can always be configured to send generic error codes.
>
> -- Christian Huitema
>

--00000000000068475005a7991941
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">I&#39;m not debating the relative merits. We let our custo=
mers set this as they wish. I&#39;m just asking what the spec is requiring =
here, and if there&#39;s an editorial change needed in one of the drafts.<d=
iv><br></div></div><br><div class=3D"gmail_quote"><div dir=3D"ltr" class=3D=
"gmail_attr">On Mon, Jun 8, 2020 at 2:12 PM Christian Huitema &lt;<a href=
=3D"mailto:huitema@huitema.net">huitema@huitema.net</a>&gt; wrote:<br></div=
><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border=
-left:1px solid rgb(204,204,204);padding-left:1ex">
 =20
   =20
 =20
  <div>
    <p>On 6/8/2020 2:00 PM, Martin Duke wrote:<br>
    </p>
    <blockquote type=3D"cite">
     =20
      <div dir=3D"ltr">In section 8.1 of quic-transport,=C2=A0
        <div><span style=3D"font-family:Lora,serif;font-size:16px">When
            using ALPN, endpoints MUST immediately close a connection
            (see Section 10.3 of=C2=A0</span><span style=3D"font-family:Lor=
a,serif;font-size:16px">[<a href=3D"https://quicwg.org/base-drafts/draft-ie=
tf-quic-tls.html#QUIC-TRANSPORT" style=3D"text-decoration-line:none;white-s=
pace:nowrap" target=3D"_blank">QUIC-TRANSPORT</a>]</span><span style=3D"fon=
t-family:Lora,serif;font-size:16px">) with a
            no_application_protocol TLS alert (QUIC error code 0x178;
            see=C2=A0</span><a href=3D"https://quicwg.org/base-drafts/draft=
-ietf-quic-tls.html#tls-errors" style=3D"text-decoration-line:none;white-sp=
ace:nowrap;font-family:Lora,serif;font-size:16px" target=3D"_blank">Section=
 4.10</a><span style=3D"font-family:Lora,serif;font-size:16px">) if an
            application protocol is not negotiated.</span>=C2=A0</div>
        <div><br>
        </div>
        <div>In 4.10 of quic-tls,</div>
        <div><span style=3D"font-family:Lora,serif;font-size:16px">Endpoint=
s
            MAY use a generic error code to avoid possibly exposing
            confidential information.</span></div>
        <div><br>
        </div>
        <div>Which one takes precedence? May the server send a generic
          error when there is no ALPN code? <br>
        </div>
      </div>
    </blockquote>
    <p><br>
    </p>
    <p>That&#39;s a very old debate. More specific the error codes make
      debugging easier but do expose details of implementation or
      configuration, so there is a tension. The usual=C2=A0 compromise is
      that privacy conscious applications can always be configured to
      send generic error codes.</p>
    <p>-- Christian Huitema<br>
    </p>
  </div>

</blockquote></div>

--00000000000068475005a7991941--


From nobody Mon Jun  8 15:29:09 2020
Return-Path: <lucaspardue.24.7@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id F01343A0869 for <quic@ietfa.amsl.com>; Mon,  8 Jun 2020 15:29:07 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.847
X-Spam-Level: 
X-Spam-Status: No, score=-1.847 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_ENVFROM_END_DIGIT=0.25, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id sue9EFQaj5K7 for <quic@ietfa.amsl.com>; Mon,  8 Jun 2020 15:29:06 -0700 (PDT)
Received: from mail-wm1-x32f.google.com (mail-wm1-x32f.google.com [IPv6:2a00:1450:4864:20::32f]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id D2B3A3A0867 for <quic@ietf.org>; Mon,  8 Jun 2020 15:29:05 -0700 (PDT)
Received: by mail-wm1-x32f.google.com with SMTP id j198so961670wmj.0 for <quic@ietf.org>; Mon, 08 Jun 2020 15:29:05 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=3Gdcj+N1TUDXylBvp4BuJMDKD03WzpU/tCbHYFWa1Bk=; b=TXvJOIKrLMgmdNoscF/0dCkTFpIRUr6ge8897Gla+ss1gWvsOUW4M1kZzFhd0L1ElH NlXWAlTiGFvOzYoVFJLrCebXg9BIzBMOcYwuwP+3ZKSHkQMHtiQGWQJF7xIAWQY+bXVl 8p0makMpsO4a8zbeKXz32Ki898qvgBX2JA4KnM+BJLt+rdY1G9wsWrnD8IthEkNZDA88 zWi01rx0GjajtUpLxmArOfy1NPZ7aeiO3n89Q4A5NJJ32aFjDNAKJ8GK1HaYQSQL3kxD yHXMGzvwYkizQD5PaqEPZwDgyVBjxR8XMMuZTz/DIxWAt2Tk1P4SUTa5kDlfV+VC0U47 /u4g==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=3Gdcj+N1TUDXylBvp4BuJMDKD03WzpU/tCbHYFWa1Bk=; b=t6gEDXn8c828eKfuVpW7gfFzCBNAidbG2SJhnSBHUp4OOKjGdKJgC3iJluFU2CcNdS /YnMJWMuQldEyqUwY30EFlnE1CCfm857k7RBlCzdGg1uzvW7KNgmANFZo1978I2NxJWh MCkzu20R1SxVMtYaTLBmrWL8qp8d03YiGPjiuReapRrruOoXBA5fwrHRxPtrjQ7cckck d675wQlddR2yYaBuh6C8xYISjaHIzp9MyheIeRIvEDbROuBjoSvW01SmIIsJmUr5/ZWS oc6FXuXtG6bMGKr0KIrJ0TlDqc8/5Uv/PUpe4T4l8EDDJWxUnzMaRjH+IMOzO8zUdwut fcLA==
X-Gm-Message-State: AOAM530H0yvm48augPb0ACP4sjAauaQkmMGiOWox1gQ2CEcYwYluesKm WPpBQvj5d/V1K4aYzVf+GsRV2VDMEfFtrnzYmV0=
X-Google-Smtp-Source: ABdhPJzVD+d5wTzBt9+e2EEEZw+220zDsSkBMP/D+w4lpFpar3OkLjnjT8df5NWf07P3DSTAqcNs+0z5FhypEJXaiqU=
X-Received: by 2002:a1c:408:: with SMTP id 8mr804717wme.15.1591655342649; Mon, 08 Jun 2020 15:29:02 -0700 (PDT)
MIME-Version: 1.0
References: <CAM4esxSg=oWNLUoNK6Rg6x1KL5NvtWz8qyvk3aa-UWL3HtaodA@mail.gmail.com> <0cde7047-7856-97c5-5e3e-0f0ad380451a@huitema.net> <CAM4esxQ+i=6d6U6iWTe5Ym=2F4-kt5BTqS-Bp=FPM1NBhnN+YA@mail.gmail.com>
In-Reply-To: <CAM4esxQ+i=6d6U6iWTe5Ym=2F4-kt5BTqS-Bp=FPM1NBhnN+YA@mail.gmail.com>
From: Lucas Pardue <lucaspardue.24.7@gmail.com>
Date: Mon, 8 Jun 2020 23:28:53 +0100
Message-ID: <CALGR9oYv58MPFYyFoPHx1a+SUtH9JJqHVOcvYmQnXF_mn69qOA@mail.gmail.com>
Subject: Re: Conflicting requirements?
To: Martin Duke <martin.h.duke@gmail.com>
Cc: Christian Huitema <huitema@huitema.net>, IETF QUIC WG <quic@ietf.org>
Content-Type: multipart/alternative; boundary="0000000000004db4d305a79a23a6"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/H_tqbt4E3pzWgV64t55IMS4IG64>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 08 Jun 2020 22:29:08 -0000

--0000000000004db4d305a79a23a6
Content-Type: text/plain; charset="UTF-8"

The intent of the language is that connection closure is MUST. The choice
of code is a bit more flexible and follows the philosophy that Christian
laid out.

On Mon, 8 Jun 2020, 22:15 Martin Duke, <martin.h.duke@gmail.com> wrote:

> I'm not debating the relative merits. We let our customers set this as
> they wish. I'm just asking what the spec is requiring here, and if there's
> an editorial change needed in one of the drafts.
>
>
> On Mon, Jun 8, 2020 at 2:12 PM Christian Huitema <huitema@huitema.net>
> wrote:
>
>> On 6/8/2020 2:00 PM, Martin Duke wrote:
>>
>> In section 8.1 of quic-transport,
>> When using ALPN, endpoints MUST immediately close a connection (see
>> Section 10.3 of [QUIC-TRANSPORT
>> <https://quicwg.org/base-drafts/draft-ietf-quic-tls.html#QUIC-TRANSPORT>])
>> with a no_application_protocol TLS alert (QUIC error code 0x178; see Section
>> 4.10 <https://quicwg.org/base-drafts/draft-ietf-quic-tls.html#tls-errors>)
>> if an application protocol is not negotiated.
>>
>> In 4.10 of quic-tls,
>> Endpoints MAY use a generic error code to avoid possibly exposing
>> confidential information.
>>
>> Which one takes precedence? May the server send a generic error when
>> there is no ALPN code?
>>
>>
>> That's a very old debate. More specific the error codes make debugging
>> easier but do expose details of implementation or configuration, so there
>> is a tension. The usual  compromise is that privacy conscious applications
>> can always be configured to send generic error codes.
>>
>> -- Christian Huitema
>>
>

--0000000000004db4d305a79a23a6
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"auto"><div>The intent of the language is that connection closur=
e is MUST. The choice of code is a bit more flexible and follows the philos=
ophy that Christian laid out.<br><br><div class=3D"gmail_quote"><div dir=3D=
"ltr" class=3D"gmail_attr">On Mon, 8 Jun 2020, 22:15 Martin Duke, &lt;<a hr=
ef=3D"mailto:martin.h.duke@gmail.com" target=3D"_blank" rel=3D"noreferrer">=
martin.h.duke@gmail.com</a>&gt; wrote:<br></div><blockquote class=3D"gmail_=
quote" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1=
ex"><div dir=3D"ltr">I&#39;m not debating the relative merits. We let our c=
ustomers set this as they wish. I&#39;m just asking what the spec is requir=
ing here, and if there&#39;s an editorial change needed in one of the draft=
s.<div><br></div></div><br><div class=3D"gmail_quote"><div dir=3D"ltr" clas=
s=3D"gmail_attr">On Mon, Jun 8, 2020 at 2:12 PM Christian Huitema &lt;<a hr=
ef=3D"mailto:huitema@huitema.net" rel=3D"noreferrer noreferrer" target=3D"_=
blank">huitema@huitema.net</a>&gt; wrote:<br></div><blockquote class=3D"gma=
il_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,2=
04,204);padding-left:1ex">
 =20
   =20
 =20
  <div>
    <p>On 6/8/2020 2:00 PM, Martin Duke wrote:<br>
    </p>
    <blockquote type=3D"cite">
     =20
      <div dir=3D"ltr">In section 8.1 of quic-transport,=C2=A0
        <div><span style=3D"font-family:Lora,serif;font-size:16px">When
            using ALPN, endpoints MUST immediately close a connection
            (see Section 10.3 of=C2=A0</span><span style=3D"font-family:Lor=
a,serif;font-size:16px">[<a href=3D"https://quicwg.org/base-drafts/draft-ie=
tf-quic-tls.html#QUIC-TRANSPORT" style=3D"text-decoration-line:none;white-s=
pace:nowrap" rel=3D"noreferrer noreferrer" target=3D"_blank">QUIC-TRANSPORT=
</a>]</span><span style=3D"font-family:Lora,serif;font-size:16px">) with a
            no_application_protocol TLS alert (QUIC error code 0x178;
            see=C2=A0</span><a href=3D"https://quicwg.org/base-drafts/draft=
-ietf-quic-tls.html#tls-errors" style=3D"text-decoration-line:none;white-sp=
ace:nowrap;font-family:Lora,serif;font-size:16px" rel=3D"noreferrer norefer=
rer" target=3D"_blank">Section 4.10</a><span style=3D"font-family:Lora,seri=
f;font-size:16px">) if an
            application protocol is not negotiated.</span>=C2=A0</div>
        <div><br>
        </div>
        <div>In 4.10 of quic-tls,</div>
        <div><span style=3D"font-family:Lora,serif;font-size:16px">Endpoint=
s
            MAY use a generic error code to avoid possibly exposing
            confidential information.</span></div>
        <div><br>
        </div>
        <div>Which one takes precedence? May the server send a generic
          error when there is no ALPN code? <br>
        </div>
      </div>
    </blockquote>
    <p><br>
    </p>
    <p>That&#39;s a very old debate. More specific the error codes make
      debugging easier but do expose details of implementation or
      configuration, so there is a tension. The usual=C2=A0 compromise is
      that privacy conscious applications can always be configured to
      send generic error codes.</p>
    <p>-- Christian Huitema<br>
    </p>
  </div>

</blockquote></div>
</blockquote></div>
</div></div>

--0000000000004db4d305a79a23a6--


From nobody Tue Jun  9 16:31:16 2020
Return-Path: <martin.h.duke@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 3859C3A0C61 for <quic@ietfa.amsl.com>; Tue,  9 Jun 2020 16:31:15 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.005
X-Spam-Level: 
X-Spam-Status: No, score=-2.005 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, HTML_OBFUSCATE_10_20=0.093, SPF_HELO_NONE=0.001, SPF_PASS=-0.001] autolearn=no autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id akgxKxbU_W5u for <quic@ietfa.amsl.com>; Tue,  9 Jun 2020 16:31:14 -0700 (PDT)
Received: from mail-il1-x12c.google.com (mail-il1-x12c.google.com [IPv6:2607:f8b0:4864:20::12c]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 1A26D3A0884 for <quic@ietf.org>; Tue,  9 Jun 2020 16:31:14 -0700 (PDT)
Received: by mail-il1-x12c.google.com with SMTP id p5so159148ile.6 for <quic@ietf.org>; Tue, 09 Jun 2020 16:31:14 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:from:date:message-id:subject:to; bh=eM0cg+Vw40WKn8gBBuZf+6gSHpObGzQ3nVL8RH0dzE8=; b=oPI/T4EfIIqS3V9ptMgazwddsKRiSDzc0HOFyNcJhLMZ4BalaUQ8fYjG866QdTQvJf mw+l5StFnTntF1oJ9jyjc2w7xai7gqpdE7zU0Pf88cuidprsg4T37rwG+6vEajl6GmyI neCY3FiPPGqhjAvtooBJ2ZyKCZ0RlXjMPoCXfG7ZaTnRxJt6PwkEasGx9WQGF1iCbtxj wxyoDpb+OUy1NuhLVAIt8/eeqUddl6pOKX5EUOvVxV+BJQ27MD35jcXlahD1JwYz6Au5 aqisr1jsx3nAQ23Whpwzt0XFxpBFL5+uma3CiR9ia/6ScAbPIzf7VE7Sfa8m2PhHbGfd KePA==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:from:date:message-id:subject:to; bh=eM0cg+Vw40WKn8gBBuZf+6gSHpObGzQ3nVL8RH0dzE8=; b=DSz5JXfxNeK8YsAaIxXPU6nbhjm0ECl0pe8vdAXQXH6AB+1e98WECX3XyUN1gysLfw idWkcw2UlcgKIIuKi4LbWyAefjyWsR7VDFePUtsD27FgGmH/Tg7R7M4+kBzGWStizqa9 HU2e1ojfiKMMK7fjTcw9qLY3sktlKXIFYxydr8aazz36Oo7ZFGPFB9Q4p7fQnt+fcxO5 EBD1S9T2oWYukVEN2dCVH+QjFIC6GRVgtRdRkCPUISuQtSQz524tM0PLkYqrNGRTmAFv GGwktBApG0/fFIojm147w6+tPpf7KWCW1TkjeN/LVyeAOW+tloL+2EuMXO7VroW2gUl1 2/Aw==
X-Gm-Message-State: AOAM533cXKF54eJSRbdsgKNcfqRqSsPELR1cXIaEvgvk4AKtr8wj8t4D 7ybwWbvIjyjmQnGjAqwLNPDtB/cKPy1zwxRLcvLK1jZtHjE=
X-Google-Smtp-Source: ABdhPJx87pwNegWIDFzCp8/BVFTXJOxpqSkcu8tbwJgE2MKQgtVky0JvEVHpOYI4HgQMVwNiZJtVN2c/8kdNSra3hCc=
X-Received: by 2002:a92:9f12:: with SMTP id u18mr391067ili.287.1591745473194;  Tue, 09 Jun 2020 16:31:13 -0700 (PDT)
MIME-Version: 1.0
From: Martin Duke <martin.h.duke@gmail.com>
Date: Tue, 9 Jun 2020 16:31:02 -0700
Message-ID: <CAM4esxTBMKH4VKt8xD4-jyamMXJBbigbQomkuC8tMp0Lu003YA@mail.gmail.com>
Subject: Version-aliasing on github
To: IETF QUIC WG <quic@ietf.org>
Content-Type: multipart/alternative; boundary="00000000000080aea005a7af1f6d"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/PicIR9g5_Rrf1ymW2HMh_xVOP4E>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 09 Jun 2020 23:31:15 -0000

--00000000000080aea005a7af1f6d
Content-Type: text/plain; charset="UTF-8"

If you're interested in this individual draft, it now has a home on github:
https://github.com/martinduke/quic-version-aliasing

--00000000000080aea005a7af1f6d
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">If you&#39;re interested=C2=A0in this individual draft, it=
 now has a home on github:<div><a href=3D"https://github.com/martinduke/qui=
c-version-aliasing">https://github.com/martinduke/quic-version-aliasing</a>=
=C2=A0</div><div>=C2=A0<br></div></div>

--00000000000080aea005a7af1f6d--


From nobody Tue Jun  9 17:00:09 2020
Return-Path: <internet-drafts@ietf.org>
X-Original-To: quic@ietf.org
Delivered-To: quic@ietfa.amsl.com
Received: from ietfa.amsl.com (localhost [IPv6:::1]) by ietfa.amsl.com (Postfix) with ESMTP id BA77D3A0C8F; Tue,  9 Jun 2020 17:00:07 -0700 (PDT)
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: 7bit
From: internet-drafts@ietf.org
To: <i-d-announce@ietf.org>
Cc: quic@ietf.org
Subject: I-D Action: draft-ietf-quic-qpack-16.txt
X-Test-IDTracker: no
X-IETF-IDTracker: 7.3.0
Auto-Submitted: auto-generated
Precedence: bulk
Reply-To: quic@ietf.org
Message-ID: <159174720769.13863.3097463932421254940@ietfa.amsl.com>
Date: Tue, 09 Jun 2020 17:00:07 -0700
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/8cXxDGfAjWw7hO7hwgADnTX-6fc>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 10 Jun 2020 00:00:08 -0000

A New Internet-Draft is available from the on-line Internet-Drafts directories.
This draft is a work item of the QUIC WG of the IETF.

        Title           : QPACK: Header Compression for HTTP/3
        Authors         : Charles 'Buck' Krasic
                          Mike Bishop
                          Alan Frindell
	Filename        : draft-ietf-quic-qpack-16.txt
	Pages           : 44
	Date            : 2020-06-09

Abstract:
   This specification defines QPACK, a compression format for
   efficiently representing HTTP fields, to be used in HTTP/3.  This is
   a variation of HPACK compression that seeks to reduce head-of-line
   blocking.


The IETF datatracker status page for this draft is:
https://datatracker.ietf.org/doc/draft-ietf-quic-qpack/

There are also htmlized versions available at:
https://tools.ietf.org/html/draft-ietf-quic-qpack-16
https://datatracker.ietf.org/doc/html/draft-ietf-quic-qpack-16

A diff from the previous version is available at:
https://www.ietf.org/rfcdiff?url2=draft-ietf-quic-qpack-16


Please note that it may take a couple of minutes from the time of submission
until the htmlized version and diff are available at tools.ietf.org.

Internet-Drafts are also available by anonymous FTP at:
ftp://ftp.ietf.org/internet-drafts/



From nobody Tue Jun  9 17:05:17 2020
Return-Path: <internet-drafts@ietf.org>
X-Original-To: quic@ietf.org
Delivered-To: quic@ietfa.amsl.com
Received: from ietfa.amsl.com (localhost [IPv6:::1]) by ietfa.amsl.com (Postfix) with ESMTP id 8F6313A0CC3; Tue,  9 Jun 2020 17:05:15 -0700 (PDT)
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: 7bit
From: internet-drafts@ietf.org
To: <i-d-announce@ietf.org>
Cc: quic@ietf.org
Subject: I-D Action: draft-ietf-quic-tls-29.txt
X-Test-IDTracker: no
X-IETF-IDTracker: 7.3.0
Auto-Submitted: auto-generated
Precedence: bulk
Reply-To: quic@ietf.org
Message-ID: <159174751554.10790.629981804498659016@ietfa.amsl.com>
Date: Tue, 09 Jun 2020 17:05:15 -0700
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/B5mn0dHKMxwqUi1c6W8Ppriab8g>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 10 Jun 2020 00:05:16 -0000

A New Internet-Draft is available from the on-line Internet-Drafts directories.
This draft is a work item of the QUIC WG of the IETF.

        Title           : Using TLS to Secure QUIC
        Authors         : Martin Thomson
                          Sean Turner
	Filename        : draft-ietf-quic-tls-29.txt
	Pages           : 59
	Date            : 2020-06-09

Abstract:
   This document describes how Transport Layer Security (TLS) is used to
   secure QUIC.

Note to Readers

   Discussion of this draft takes place on the QUIC working group
   mailing list (quic@ietf.org (mailto:quic@ietf.org)), which is
   archived at https://mailarchive.ietf.org/arch/
   search/?email_list=quic.

   Working Group information can be found at https://github.com/quicwg;
   source code and issues list for this draft can be found at
   https://github.com/quicwg/base-drafts/labels/-tls.


The IETF datatracker status page for this draft is:
https://datatracker.ietf.org/doc/draft-ietf-quic-tls/

There are also htmlized versions available at:
https://tools.ietf.org/html/draft-ietf-quic-tls-29
https://datatracker.ietf.org/doc/html/draft-ietf-quic-tls-29

A diff from the previous version is available at:
https://www.ietf.org/rfcdiff?url2=draft-ietf-quic-tls-29


Please note that it may take a couple of minutes from the time of submission
until the htmlized version and diff are available at tools.ietf.org.

Internet-Drafts are also available by anonymous FTP at:
ftp://ftp.ietf.org/internet-drafts/



From nobody Tue Jun  9 17:05:23 2020
Return-Path: <internet-drafts@ietf.org>
X-Original-To: quic@ietf.org
Delivered-To: quic@ietfa.amsl.com
Received: from ietfa.amsl.com (localhost [IPv6:::1]) by ietfa.amsl.com (Postfix) with ESMTP id 1909A3A053F; Tue,  9 Jun 2020 17:05:16 -0700 (PDT)
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: 7bit
From: internet-drafts@ietf.org
To: <i-d-announce@ietf.org>
Cc: quic@ietf.org
Subject: I-D Action: draft-ietf-quic-invariants-09.txt
X-Test-IDTracker: no
X-IETF-IDTracker: 7.3.0
Auto-Submitted: auto-generated
Precedence: bulk
Reply-To: quic@ietf.org
Message-ID: <159174751606.15484.11263196257628535808@ietfa.amsl.com>
Date: Tue, 09 Jun 2020 17:05:16 -0700
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/tA1cYyCcNct6VvJ6JWTqeP1pR8E>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 10 Jun 2020 00:05:16 -0000

A New Internet-Draft is available from the on-line Internet-Drafts directories.
This draft is a work item of the QUIC WG of the IETF.

        Title           : Version-Independent Properties of QUIC
        Author          : Martin Thomson
	Filename        : draft-ietf-quic-invariants-09.txt
	Pages           : 10
	Date            : 2020-06-09

Abstract:
   This document defines the properties of the QUIC transport protocol
   that are expected to remain unchanged over time as new versions of
   the protocol are developed.


The IETF datatracker status page for this draft is:
https://datatracker.ietf.org/doc/draft-ietf-quic-invariants/

There are also htmlized versions available at:
https://tools.ietf.org/html/draft-ietf-quic-invariants-09
https://datatracker.ietf.org/doc/html/draft-ietf-quic-invariants-09

A diff from the previous version is available at:
https://www.ietf.org/rfcdiff?url2=draft-ietf-quic-invariants-09


Please note that it may take a couple of minutes from the time of submission
until the htmlized version and diff are available at tools.ietf.org.

Internet-Drafts are also available by anonymous FTP at:
ftp://ftp.ietf.org/internet-drafts/



From nobody Tue Jun  9 17:07:12 2020
Return-Path: <internet-drafts@ietf.org>
X-Original-To: quic@ietf.org
Delivered-To: quic@ietfa.amsl.com
Received: from ietfa.amsl.com (localhost [IPv6:::1]) by ietfa.amsl.com (Postfix) with ESMTP id D30E93A0CEE; Tue,  9 Jun 2020 17:07:11 -0700 (PDT)
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: 7bit
From: internet-drafts@ietf.org
To: <i-d-announce@ietf.org>
Cc: quic@ietf.org
Subject: I-D Action: draft-ietf-quic-transport-29.txt
X-Test-IDTracker: no
X-IETF-IDTracker: 7.3.0
Auto-Submitted: auto-generated
Precedence: bulk
Reply-To: quic@ietf.org
Message-ID: <159174763182.12616.9533113763555229166@ietfa.amsl.com>
Date: Tue, 09 Jun 2020 17:07:11 -0700
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/gGE_K29GBAzXPm47oEQDPp4zL9Y>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 10 Jun 2020 00:07:12 -0000

A New Internet-Draft is available from the on-line Internet-Drafts directories.
This draft is a work item of the QUIC WG of the IETF.

        Title           : QUIC: A UDP-Based Multiplexed and Secure Transport
        Authors         : Jana Iyengar
                          Martin Thomson
	Filename        : draft-ietf-quic-transport-29.txt
	Pages           : 187
	Date            : 2020-06-09

Abstract:
   This document defines the core of the QUIC transport protocol.
   Accompanying documents describe QUIC's loss detection and congestion
   control and the use of TLS for key negotiation.

Note to Readers

   Discussion of this draft takes place on the QUIC working group
   mailing list (quic@ietf.org (mailto:quic@ietf.org)), which is
   archived at https://mailarchive.ietf.org/arch/search/?email_list=quic

   Working Group information can be found at https://github.com/quicwg;
   source code and issues list for this draft can be found at
   https://github.com/quicwg/base-drafts/labels/-transport.


The IETF datatracker status page for this draft is:
https://datatracker.ietf.org/doc/draft-ietf-quic-transport/

There are also htmlized versions available at:
https://tools.ietf.org/html/draft-ietf-quic-transport-29
https://datatracker.ietf.org/doc/html/draft-ietf-quic-transport-29

A diff from the previous version is available at:
https://www.ietf.org/rfcdiff?url2=draft-ietf-quic-transport-29


Please note that it may take a couple of minutes from the time of submission
until the htmlized version and diff are available at tools.ietf.org.

Internet-Drafts are also available by anonymous FTP at:
ftp://ftp.ietf.org/internet-drafts/



From nobody Tue Jun  9 17:34:30 2020
Return-Path: <internet-drafts@ietf.org>
X-Original-To: quic@ietf.org
Delivered-To: quic@ietfa.amsl.com
Received: from ietfa.amsl.com (localhost [IPv6:::1]) by ietfa.amsl.com (Postfix) with ESMTP id 248793A0CE0; Tue,  9 Jun 2020 17:34:29 -0700 (PDT)
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: 7bit
From: internet-drafts@ietf.org
To: <i-d-announce@ietf.org>
Cc: quic@ietf.org
Subject: I-D Action: draft-ietf-quic-recovery-29.txt
X-Test-IDTracker: no
X-IETF-IDTracker: 7.3.0
Auto-Submitted: auto-generated
Precedence: bulk
Reply-To: quic@ietf.org
Message-ID: <159174926905.11646.1975231547639763889@ietfa.amsl.com>
Date: Tue, 09 Jun 2020 17:34:29 -0700
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/CsH4SS0CyiozV15tzB0qZFwIlkU>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 10 Jun 2020 00:34:29 -0000

A New Internet-Draft is available from the on-line Internet-Drafts directories.
This draft is a work item of the QUIC WG of the IETF.

        Title           : QUIC Loss Detection and Congestion Control
        Authors         : Jana Iyengar
                          Ian Swett
	Filename        : draft-ietf-quic-recovery-29.txt
	Pages           : 46
	Date            : 2020-06-09

Abstract:
   This document describes loss detection and congestion control
   mechanisms for QUIC.

Note to Readers

   Discussion of this draft takes place on the QUIC working group
   mailing list (quic@ietf.org (mailto:quic@ietf.org)), which is
   archived at https://mailarchive.ietf.org/arch/
   search/?email_list=quic.

   Working Group information can be found at https://github.com/quicwg;
   source code and issues list for this draft can be found at
   https://github.com/quicwg/base-drafts/labels/-recovery.


The IETF datatracker status page for this draft is:
https://datatracker.ietf.org/doc/draft-ietf-quic-recovery/

There are also htmlized versions available at:
https://tools.ietf.org/html/draft-ietf-quic-recovery-29
https://datatracker.ietf.org/doc/html/draft-ietf-quic-recovery-29

A diff from the previous version is available at:
https://www.ietf.org/rfcdiff?url2=draft-ietf-quic-recovery-29


Please note that it may take a couple of minutes from the time of submission
until the htmlized version and diff are available at tools.ietf.org.

Internet-Drafts are also available by anonymous FTP at:
ftp://ftp.ietf.org/internet-drafts/



From nobody Tue Jun  9 17:50:49 2020
Return-Path: <internet-drafts@ietf.org>
X-Original-To: quic@ietf.org
Delivered-To: quic@ietfa.amsl.com
Received: from ietfa.amsl.com (localhost [IPv6:::1]) by ietfa.amsl.com (Postfix) with ESMTP id 3387B3A0D00; Tue,  9 Jun 2020 17:50:47 -0700 (PDT)
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: 7bit
From: internet-drafts@ietf.org
To: <i-d-announce@ietf.org>
Cc: quic@ietf.org
Subject: I-D Action: draft-ietf-quic-http-29.txt
X-Test-IDTracker: no
X-IETF-IDTracker: 7.3.0
Auto-Submitted: auto-generated
Precedence: bulk
Reply-To: quic@ietf.org
Message-ID: <159175024717.8667.7247432328847017660@ietfa.amsl.com>
Date: Tue, 09 Jun 2020 17:50:47 -0700
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/AdgXF832LXVEI6eaAP-4YnS95Sk>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 10 Jun 2020 00:50:47 -0000

A New Internet-Draft is available from the on-line Internet-Drafts directories.
This draft is a work item of the QUIC WG of the IETF.

        Title           : Hypertext Transfer Protocol Version 3 (HTTP/3)
        Author          : Mike Bishop
	Filename        : draft-ietf-quic-http-29.txt
	Pages           : 72
	Date            : 2020-06-09

Abstract:
   The QUIC transport protocol has several features that are desirable
   in a transport for HTTP, such as stream multiplexing, per-stream flow
   control, and low-latency connection establishment.  This document
   describes a mapping of HTTP semantics over QUIC.  This document also
   identifies HTTP/2 features that are subsumed by QUIC, and describes
   how HTTP/2 extensions can be ported to HTTP/3.


The IETF datatracker status page for this draft is:
https://datatracker.ietf.org/doc/draft-ietf-quic-http/

There are also htmlized versions available at:
https://tools.ietf.org/html/draft-ietf-quic-http-29
https://datatracker.ietf.org/doc/html/draft-ietf-quic-http-29

A diff from the previous version is available at:
https://www.ietf.org/rfcdiff?url2=draft-ietf-quic-http-29


Please note that it may take a couple of minutes from the time of submission
until the htmlized version and diff are available at tools.ietf.org.

Internet-Drafts are also available by anonymous FTP at:
ftp://ftp.ietf.org/internet-drafts/



From nobody Tue Jun  9 18:36:20 2020
Return-Path: <lucaspardue.24.7@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 6A96C3A0D6F for <quic@ietfa.amsl.com>; Tue,  9 Jun 2020 18:36:18 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.847
X-Spam-Level: 
X-Spam-Status: No, score=-1.847 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_ENVFROM_END_DIGIT=0.25, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 6TcHB4IAB4aB for <quic@ietfa.amsl.com>; Tue,  9 Jun 2020 18:36:16 -0700 (PDT)
Received: from mail-wr1-x430.google.com (mail-wr1-x430.google.com [IPv6:2a00:1450:4864:20::430]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 6E8E13A0D6D for <quic@ietf.org>; Tue,  9 Jun 2020 18:36:16 -0700 (PDT)
Received: by mail-wr1-x430.google.com with SMTP id l11so405485wru.0 for <quic@ietf.org>; Tue, 09 Jun 2020 18:36:16 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:from:date:message-id:subject:to; bh=CAbgDTnjLmYDqfHnId5O44Ktjo/tWLY4blPVaflT+IE=; b=sNgY2IUArgQfJ4atwPqEeRHZjjy3ro+lqU0bglaHxM9xbYYVOu9M4RlgizmDBpzSgs ib/ymftNyrCJfQaSBpgNNdOx08AYk3MHdXOcFjufiiUjMANNo1kTsKJlqXQGnRXsnYZO o55bfM9DsrAV439xyHbdjGxg9YKxhS/8z29qLwIyc7Z2AdUgTqo9xkJIxRSDRDOTCKM3 2zKyqLapRDzutIAo8+W+mbnYlYkGIht2B2jPsjTt62eyNzu+68fj8DGK/7u7rkc9itnZ Va7nHg7luBB9raIogrNhOeyEHwlS0FuxO3CMSuEj5UzBdigYjBptZ1VvD4rJbBK6cmyj pF2w==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:from:date:message-id:subject:to; bh=CAbgDTnjLmYDqfHnId5O44Ktjo/tWLY4blPVaflT+IE=; b=XGVNeD+fEUClOFCkIxcG9P3/g1sZxyrFY2ErgnzZZu2dStfpjh/pPIWjVzE5pau1Rv 1fsN2/2ozLdk1vi9L3fsoD+iAGreicyjg3laOTWxJ4CcYZPcOmgh36gFhf3/vhnckkoG WwggRtk4qa/bZsW8OQjp0pp627Ezi8YTx0enIIhznQZ9PuIEXyVzQvxVGBoQZpDjCKSg hWsVnyYOvmLXbdonVceoMAOHVkOTYY7NFp+FiTdmrgxUTq+wCebRnz/GQd3Z3VViW6Q2 zDx5++YrTmwRvzWBVna4ATCaF6nurazQejCjW7eJ8wI4RrszdQy574aOrqbIn4Epw1so dtJA==
X-Gm-Message-State: AOAM531IQQQloubGKMj9N1CQhI/rQOZntBD6muM1qKtp90IpaaXtjx/L 5dVLqWFnvmsCw5QkO8uOwsM6+VqwdvyC4LiQGs89xs6/
X-Google-Smtp-Source: ABdhPJxUX7gcynKxnFW8ohy3PTCxY0MPBuKYZWKZouyOYjr8h59tqKBZLTAT7+Dylrz/pAUo4Pn1hQ7YhQAOmoVeRiU=
X-Received: by 2002:a5d:4008:: with SMTP id n8mr749054wrp.82.1591752973187; Tue, 09 Jun 2020 18:36:13 -0700 (PDT)
MIME-Version: 1.0
From: Lucas Pardue <lucaspardue.24.7@gmail.com>
Date: Wed, 10 Jun 2020 02:36:02 +0100
Message-ID: <CALGR9oZ9u=huobnQig0mMPS=-Fu7Mu3q8GHLTBOxd2W5u0h_kw@mail.gmail.com>
Subject: Working Group Last Call: QUIC protocol drafts
To: QUIC WG <quic@ietf.org>
Content-Type: multipart/alternative; boundary="000000000000897ebd05a7b0de1b"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/F7wvKGnA1FJasmaE35XIxsc2Tno>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 10 Jun 2020 01:36:18 -0000

--000000000000897ebd05a7b0de1b
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Hello,

After more than three and a half years and substantial discussion, all 845
of the design issues raised against the QUIC protocol drafts have gained
consensus or have a proposed resolution. In that time the protocol has been
considerably transformed; it has become more secure, much more widely
implemented, and has been shown to be interoperable. Both the Chairs and
the Editors feel that it is ready to proceed in standardisation.

Therefore, this email announces a Working Group Last Call (WGLC) for the
following QUIC documents:

* QUIC Transport
  https://tools.ietf.org/html/draft-ietf-quic-transport-29

* QUIC Loss Detection and Congestion Control
  https://tools.ietf.org/html/draft-ietf-quic-recovery-29

* Using TLS to Secure QUIC
  https://tools.ietf.org/html/draft-ietf-quic-tls-29

* Version-Independent Properties of QUIC
  https://tools.ietf.org/html/draft-ietf-quic-invariants-09

* HTTP/3
  https://tools.ietf.org/html/draft-ietf-quic-http-29

* QPACK Header Compression for HTTP/3
  https://tools.ietf.org/html/draft-ietf-quic-qpack-16

The WGLC will run for four weeks, ending on 8 July 2020.

As a reminder, we have been operating under the Late-Stage Process; see
https://github.com/quicwg/base-drafts/blob/master/CONTRIBUTING.md#late-stag=
e-process.
In theory, this means that the contents of the drafts above already have
consensus. However, the Chairs would like to actively reaffirm that
consensus and start the process of wider review through a formal WGLC.

Please review the documents above and open issues for your review comments
in our repository at https://github.com/quicwg/base-drafts. You may also
send comments to quic@ietf.org.

Issues raised during WGLC will be handled in accordance with the Late-Stage
Process defined in the Contribution Guidelines (see link above). Please
note that design issues that revisit a topic where there's already declared
consensus (see
https://github.com/quicwg/base-drafts/issues?q=3Dis%3Aclosed+label%3Ahas-co=
nsensus)
need to provide compelling reasons to warrant reopening the discussion.

As part of this WGLC, we seek consensus on the remaining open design issue
#3661 =E2=80=9CInclude epoch in the AAD or the nonce?=E2=80=9D (
https://github.com/quicwg/base-drafts/issues/3661). The proposed resolution
for this issue is to close with no action, which means that the drafts
above already reflect this emerging consensus.

Subject to the feedback received during this WGLC, a subsequent smaller
WGLC may be run in the near future to confirm any changes to the drafts
made between now and then.

The Applicability and Manageability drafts have some dependencies on the
core drafts, so we'll run separate WGLCs for them.

Cheers,

Lars, Lucas and Mark
QUIC WG Chairs

--000000000000897ebd05a7b0de1b
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Hello,<br><br>After more than three and a half years and s=
ubstantial discussion, all 845 of the design issues raised against the QUIC=
 protocol drafts have gained consensus or have a proposed resolution. In th=
at time the protocol has been considerably transformed; it has become more =
secure, much more widely implemented, and has been shown to be interoperabl=
e. Both the Chairs and the Editors feel that it is ready to proceed in stan=
dardisation.<br><br>Therefore, this email announces a Working Group Last Ca=
ll (WGLC) for the following QUIC documents:<br><br>* QUIC Transport<br>=C2=
=A0 <a href=3D"https://tools.ietf.org/html/draft-ietf-quic-transport-29">ht=
tps://tools.ietf.org/html/draft-ietf-quic-transport-29</a><br><br>* QUIC Lo=
ss Detection and Congestion Control<br>=C2=A0 <a href=3D"https://tools.ietf=
.org/html/draft-ietf-quic-recovery-29">https://tools.ietf.org/html/draft-ie=
tf-quic-recovery-29</a><br><br>* Using TLS to Secure QUIC<br>=C2=A0 <a href=
=3D"https://tools.ietf.org/html/draft-ietf-quic-tls-29">https://tools.ietf.=
org/html/draft-ietf-quic-tls-29</a><br><br>* Version-Independent Properties=
 of QUIC<br>=C2=A0 <a href=3D"https://tools.ietf.org/html/draft-ietf-quic-i=
nvariants-09">https://tools.ietf.org/html/draft-ietf-quic-invariants-09</a>=
<br><br>* HTTP/3<br>=C2=A0 <a href=3D"https://tools.ietf.org/html/draft-iet=
f-quic-http-29">https://tools.ietf.org/html/draft-ietf-quic-http-29</a><br>=
<br>* QPACK Header Compression for HTTP/3<br>=C2=A0 <a href=3D"https://tool=
s.ietf.org/html/draft-ietf-quic-qpack-16">https://tools.ietf.org/html/draft=
-ietf-quic-qpack-16</a><br><br>The WGLC will run for four weeks, ending on =
8 July 2020.<br><br>As a reminder, we have been operating under the Late-St=
age Process; see <a href=3D"https://github.com/quicwg/base-drafts/blob/mast=
er/CONTRIBUTING.md#late-stage-process">https://github.com/quicwg/base-draft=
s/blob/master/CONTRIBUTING.md#late-stage-process</a><span class=3D"sewm0wbm=
768v0q9"></span><span class=3D"sew3618dartkzuo"></span>. In theory, this me=
ans that the contents of the drafts above already have consensus. However, =
the Chairs would like to actively reaffirm that consensus and start the pro=
cess of wider review through a formal WGLC.<br><br>Please review the docume=
nts above and open issues for your review comments in our repository at <a =
href=3D"https://github.com/quicwg/base-drafts">https://github.com/quicwg/ba=
se-drafts</a>. You may also send comments to <a href=3D"mailto:quic@ietf.or=
g">quic@ietf.org</a>.<br><br>Issues raised during WGLC will be handled in a=
ccordance with the Late-Stage Process defined in the Contribution Guideline=
s (see link above). Please note that design issues that revisit a topic whe=
re there&#39;s already declared consensus (see <a href=3D"https://github.co=
m/quicwg/base-drafts/issues?q=3Dis%3Aclosed+label%3Ahas-consensus">https://=
github.com/quicwg/base-drafts/issues?q=3Dis%3Aclosed+label%3Ahas-consensus<=
/a>) need to provide compelling reasons to warrant reopening the discussion=
.<br><br>As part of this WGLC, we seek consensus on the remaining open desi=
gn issue #3661 =E2=80=9CInclude epoch in the AAD or the nonce?=E2=80=9D (<a=
 href=3D"https://github.com/quicwg/base-drafts/issues/3661">https://github.=
com/quicwg/base-drafts/issues/3661</a>). The proposed resolution for this i=
ssue is to close with no action, which means that the drafts above already =
reflect this emerging consensus.<br><br>Subject to the feedback received du=
ring this WGLC, a subsequent smaller WGLC may be run in the near future to =
confirm any changes to the drafts made between now and then.<br><br>The App=
licability and Manageability drafts have some dependencies on the core draf=
ts, so we&#39;ll run separate WGLCs for them.<br><br>Cheers,<br><br>Lars, L=
ucas and Mark<br>QUIC WG Chairs<br></div>

--000000000000897ebd05a7b0de1b--


From nobody Wed Jun 10 01:08:25 2020
Return-Path: <lars@eggert.org>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 28C213A0F9B for <quic@ietfa.amsl.com>; Wed, 10 Jun 2020 01:08:23 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.098
X-Spam-Level: 
X-Spam-Status: No, score=-2.098 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=eggert.org
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id atrrmmeaRCxu for <quic@ietfa.amsl.com>; Wed, 10 Jun 2020 01:08:21 -0700 (PDT)
Received: from mail.eggert.org (mail.eggert.org [IPv6:2a00:ac00:0:35:211:32ff:fe22:186f]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 0C1413A0F93 for <quic@ietf.org>; Wed, 10 Jun 2020 01:08:21 -0700 (PDT)
Received: from [IPv6:2a00:ac00:0:35:9050:d57a:74ca:d3ea] (unknown [IPv6:2a00:ac00:0:35:9050:d57a:74ca:d3ea]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.eggert.org (Postfix) with ESMTPSA id F222B61CC14 for <quic@ietf.org>; Wed, 10 Jun 2020 11:07:54 +0300 (EEST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=eggert.org; s=dkim; t=1591776475; bh=YfuutAIZYi3nCD2ujhMzheVVJdTSFKw3ogShMW6eCho=; h=From:Subject:References:To:Date; b=e6sdwUb3TyXpOhsR8IzYGbwjSinSj4AwSPSNlgQAtL4bquYvUP8/ovQLZdET6BYvp Ocz7jr7sHvHhfWhj2bYfKlAFiHutXleFuroNLMeRdQRKHKQQ4DG6LjhY7H+Ds0Vz/0 zpOrx1Uvrwcb6jHKsUslWHu8IdsuX2OrFz1IY/s4=
From: Lars Eggert <lars@eggert.org>
Content-Type: multipart/signed; boundary="Apple-Mail=_0047D1F0-699F-428F-932A-6947D02C370C"; protocol="application/pgp-signature"; micalg=pgp-sha512
Mime-Version: 1.0 (Mac OS X Mail 13.4 \(3608.80.23.2.2\))
Subject: Fwd: [tsvwg] 3rd WGLC (limited-scope): draft-ietf-tsvwg-transport-encrypt-15, closes 29 June 2020
Message-Id: <63DFB8B9-83DA-445E-AB71-1486D7BA33B4@eggert.org>
References: <MN2PR19MB40450EE357BEECD723AB06F183820@MN2PR19MB4045.namprd19.prod.outlook.com>
To: IETF QUIC WG <quic@ietf.org>
Date: Wed, 10 Jun 2020 11:07:51 +0300
X-MailScanner-ID: F222B61CC14.A6E12
X-MailScanner: Found to be clean
X-MailScanner-From: lars@eggert.org
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/atXg9HJdxZXxfKGlS48NdV9MXKY>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 10 Jun 2020 08:08:23 -0000

--Apple-Mail=_0047D1F0-699F-428F-932A-6947D02C370C
Content-Type: multipart/alternative;
	boundary="Apple-Mail=_07F27AFB-2557-4BD3-A8FF-4AE38A9E3021"


--Apple-Mail=_07F27AFB-2557-4BD3-A8FF-4AE38A9E3021
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=utf-8

FYI

> Begin forwarded message:
>=20
> From: "Black, David" <David.Black@dell.com>
> Subject: [tsvwg] 3rd WGLC (limited-scope): =
draft-ietf-tsvwg-transport-encrypt-15, closes 29 June 2020
> Date: June 9, 2020 at 4:41:40 GMT+3
> To: "tsvwg@ietf.org" <tsvwg@ietf.org>
> Cc: int-area <int-area@ietf.org>, IETF SAAG <saag@ietf.org>
>=20
> This email announces a limited-scope 3rd TSVWG Working Group Last Call =
(WGLC) on:
>=20
>     Considerations around Transport Header Confidentiality, Network
>      Operations, and the Evolution of Internet Transport Protocols
>                  draft-ietf-tsvwg-transport-encrypt-15
> https://datatracker.ietf.org/doc/draft-ietf-tsvwg-transport-encrypt/ =
<https://datatracker.ietf.org/doc/draft-ietf-tsvwg-transport-encrypt/>
>=20
> This draft is intended to become an Informational RFC.  This WGLC has
> been cc:=E2=80=99d to the SAAG and INT-AREA lists courtesy of the =
breadth of
> interest in this draft, but WGLC discussion will take place on the =
TSVWG
> list (tsvwg@ietf.org <mailto:tsvwg@ietf.org>) =E2=80=93 please don=E2=80=
=99t remove that list address if/when
> replying with WGLC comments.
>=20
> This 3rd WGLC will run through the end of the day on Monday, June 29,
> 2 weeks before the draft submission cutoff for IETF 108.
>=20
> This 3rd WGLC is limited to the following two topics:
>=20
> Whether or not to proceed with a request for RFC publication
> of the draft.   The decision on whether or not to proceed will
> be based on rough consensus of the WG, see RFC 7282.
> During the 2nd WGLC, Eric Rescorla and David Schinazi expressed
> strong views that this draft should not be published =E2=80=93  those
> concerns have not been resolved and are carried forward to
> this WGLC.  This email message was an attempt to summarize
> those concerns:
> =
https://mailarchive.ietf.org/arch/msg/tsvwg/i4qyY1HRqKwm0Jme9UtEb6DyhXU/ =
<https://mailarchive.ietf.org/arch/msg/tsvwg/i4qyY1HRqKwm0Jme9UtEb6DyhXU/>=

> Further explanation from both Eric Rescorla and David Schinazi
> is welcome and encouraged to ensure that their concerns are
> clearly understood.
>=20
> Review of changes made since the -12 version of the draft that
> was the subject of the second WGLC (e.g., whether or not they
> suffice to resolve concerns raised during that WGLC, other
> than overall objections to publishing this draft as an RFC):
> =
https://www.ietf.org/rfcdiff?url1=3Ddraft-ietf-tsvwg-transport-encrypt-12&=
url2=3Ddraft-ietf-tsvwg-transport-encrypt-15 =
<https://www.ietf.org/rfcdiff?url1=3Ddraft-ietf-tsvwg-transport-encrypt-12=
&url2=3Ddraft-ietf-tsvwg-transport-encrypt-15>
>=20
> Comments should be sent to the tsvwg@ietf.org <mailto:tsvwg@ietf.org> =
list, although purely
> editorial comments may be sent directly to the authors.  Please cc: =
the
> WG chairs at tsvwg-chairs@ietf.org <mailto:tsvwg-chairs@ietf.org>  if =
you would like the chairs to
> track such editorial comments as part of the WGLC process.
>=20
> No IPR disclosures have been submitted directly on this draft.
>=20
> Thanks,
> David and Wes (TSVWG Co-Chairs =E2=80=93 Gorry is recused as a draft =
author)


--Apple-Mail=_07F27AFB-2557-4BD3-A8FF-4AE38A9E3021
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html;
	charset=utf-8

<html><head><meta http-equiv=3D"Content-Type" content=3D"text/html; =
charset=3Dutf-8"></head><body style=3D"word-wrap: break-word; =
-webkit-nbsp-mode: space; line-break: after-white-space;" =
class=3D"">FYI<br class=3D""><div><br class=3D""><blockquote type=3D"cite"=
 class=3D""><div class=3D"">Begin forwarded message:</div><br =
class=3D"Apple-interchange-newline"><div style=3D"margin-top: 0px; =
margin-right: 0px; margin-bottom: 0px; margin-left: 0px;" class=3D""><span=
 style=3D"font-family: -webkit-system-font, Helvetica Neue, Helvetica, =
sans-serif; color:rgba(0, 0, 0, 1.0);" class=3D""><b class=3D"">From: =
</b></span><span style=3D"font-family: -webkit-system-font, Helvetica =
Neue, Helvetica, sans-serif;" class=3D"">"Black, David" &lt;<a =
href=3D"mailto:David.Black@dell.com" =
class=3D"">David.Black@dell.com</a>&gt;<br class=3D""></span></div><div =
style=3D"margin-top: 0px; margin-right: 0px; margin-bottom: 0px; =
margin-left: 0px;" class=3D""><span style=3D"font-family: =
-webkit-system-font, Helvetica Neue, Helvetica, sans-serif; =
color:rgba(0, 0, 0, 1.0);" class=3D""><b class=3D"">Subject: =
</b></span><span style=3D"font-family: -webkit-system-font, Helvetica =
Neue, Helvetica, sans-serif;" class=3D""><b class=3D"">[tsvwg] 3rd WGLC =
(limited-scope): draft-ietf-tsvwg-transport-encrypt-15, closes 29 June =
2020</b><br class=3D""></span></div><div style=3D"margin-top: 0px; =
margin-right: 0px; margin-bottom: 0px; margin-left: 0px;" class=3D""><span=
 style=3D"font-family: -webkit-system-font, Helvetica Neue, Helvetica, =
sans-serif; color:rgba(0, 0, 0, 1.0);" class=3D""><b class=3D"">Date: =
</b></span><span style=3D"font-family: -webkit-system-font, Helvetica =
Neue, Helvetica, sans-serif;" class=3D"">June 9, 2020 at 4:41:40 =
GMT+3<br class=3D""></span></div><div style=3D"margin-top: 0px; =
margin-right: 0px; margin-bottom: 0px; margin-left: 0px;" class=3D""><span=
 style=3D"font-family: -webkit-system-font, Helvetica Neue, Helvetica, =
sans-serif; color:rgba(0, 0, 0, 1.0);" class=3D""><b class=3D"">To: =
</b></span><span style=3D"font-family: -webkit-system-font, Helvetica =
Neue, Helvetica, sans-serif;" class=3D"">"<a =
href=3D"mailto:tsvwg@ietf.org" class=3D"">tsvwg@ietf.org</a>" &lt;<a =
href=3D"mailto:tsvwg@ietf.org" class=3D"">tsvwg@ietf.org</a>&gt;<br =
class=3D""></span></div><div style=3D"margin-top: 0px; margin-right: =
0px; margin-bottom: 0px; margin-left: 0px;" class=3D""><span =
style=3D"font-family: -webkit-system-font, Helvetica Neue, Helvetica, =
sans-serif; color:rgba(0, 0, 0, 1.0);" class=3D""><b class=3D"">Cc: =
</b></span><span style=3D"font-family: -webkit-system-font, Helvetica =
Neue, Helvetica, sans-serif;" class=3D"">int-area &lt;<a =
href=3D"mailto:int-area@ietf.org" class=3D"">int-area@ietf.org</a>&gt;, =
IETF SAAG &lt;<a href=3D"mailto:saag@ietf.org" =
class=3D"">saag@ietf.org</a>&gt;<br class=3D""></span></div><br =
class=3D""><div class=3D""><div class=3D"WordSection1" style=3D"page: =
WordSection1; caret-color: rgb(0, 0, 0); font-family: Menlo-Regular; =
font-size: 12px; font-style: normal; font-variant-caps: normal; =
font-weight: normal; letter-spacing: normal; text-align: start; =
text-indent: 0px; text-transform: none; white-space: normal; =
word-spacing: 0px; -webkit-text-stroke-width: 0px; text-decoration: =
none;"><div style=3D"margin: 0in 0in 0.0001pt; font-size: 11pt; =
font-family: Calibri, sans-serif;" class=3D"">This email announces a =
limited-scope 3rd TSVWG Working Group Last Call (WGLC) on:<o:p =
class=3D""></o:p></div><div style=3D"margin: 0in 0in 0.0001pt; =
font-size: 11pt; font-family: Calibri, sans-serif;" class=3D""><o:p =
class=3D"">&nbsp;</o:p></div><div style=3D"margin: 0in 0in 0.0001pt; =
font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D"">&nbsp;&nbsp;&nbsp; Considerations around Transport Header =
Confidentiality, Network<o:p class=3D""></o:p></div><div style=3D"margin: =
0in 0in 0.0001pt; font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D"">&nbsp;&nbsp;&nbsp;&nbsp; Operations, and the Evolution of =
Internet Transport Protocols<o:p class=3D""></o:p></div><div =
style=3D"margin: 0in 0in 0.0001pt; font-size: 11pt; font-family: =
Calibri, sans-serif;" =
class=3D"">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; =
draft-ietf-tsvwg-transport-encrypt-15<o:p class=3D""></o:p></div><div =
style=3D"margin: 0in 0in 0.0001pt; font-size: 11pt; font-family: =
Calibri, sans-serif;" class=3D""><a =
href=3D"https://datatracker.ietf.org/doc/draft-ietf-tsvwg-transport-encryp=
t/" style=3D"color: rgb(149, 79, 114); text-decoration: underline;" =
class=3D"">https://datatracker.ietf.org/doc/draft-ietf-tsvwg-transport-enc=
rypt/</a><o:p class=3D""></o:p></div><div style=3D"margin: 0in 0in =
0.0001pt; font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D""><o:p class=3D"">&nbsp;</o:p></div><div style=3D"margin: 0in =
0in 0.0001pt; font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D"">This draft is intended to become an Informational RFC.&nbsp; =
This WGLC has<o:p class=3D""></o:p></div><div style=3D"margin: 0in 0in =
0.0001pt; font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D"">been cc:=E2=80=99d to the SAAG and INT-AREA lists courtesy of =
the breadth of<o:p class=3D""></o:p></div><div style=3D"margin: 0in 0in =
0.0001pt; font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D"">interest in this draft, but WGLC discussion will take place =
on the TSVWG<o:p class=3D""></o:p></div><div style=3D"margin: 0in 0in =
0.0001pt; font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D"">list (<a href=3D"mailto:tsvwg@ietf.org" style=3D"color: =
rgb(149, 79, 114); text-decoration: underline;" =
class=3D"">tsvwg@ietf.org</a>) =E2=80=93 please don=E2=80=99t remove =
that list address if/when<o:p class=3D""></o:p></div><div style=3D"margin:=
 0in 0in 0.0001pt; font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D"">replying with WGLC comments.<o:p class=3D""></o:p></div><div =
style=3D"margin: 0in 0in 0.0001pt; font-size: 11pt; font-family: =
Calibri, sans-serif;" class=3D""><o:p class=3D"">&nbsp;</o:p></div><div =
style=3D"margin: 0in 0in 0.0001pt; font-size: 11pt; font-family: =
Calibri, sans-serif;" class=3D"">This 3rd WGLC will run through the end =
of the day on Monday, June 29,<o:p class=3D""></o:p></div><div =
style=3D"margin: 0in 0in 0.0001pt; font-size: 11pt; font-family: =
Calibri, sans-serif;" class=3D"">2 weeks before the draft submission =
cutoff for IETF 108.<o:p class=3D""></o:p></div><div style=3D"margin: =
0in 0in 0.0001pt; font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D""><o:p class=3D"">&nbsp;</o:p></div><div style=3D"margin: 0in =
0in 0.0001pt; font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D"">This 3rd WGLC is limited to the following two topics:<o:p =
class=3D""></o:p></div><div style=3D"margin: 0in 0in 0.0001pt; =
font-size: 11pt; font-family: Calibri, sans-serif;" class=3D""><o:p =
class=3D"">&nbsp;</o:p></div><ol start=3D"1" type=3D"1" =
style=3D"margin-bottom: 0in; margin-top: 0in;" class=3D""><li =
class=3D"MsoPlainText" style=3D"margin: 0in 0in 0.0001pt; font-size: =
11pt; font-family: Calibri, sans-serif;">Whether or not to proceed with =
a request for RFC publication<o:p class=3D""></o:p></li></ol><div =
style=3D"margin: 0in 0in 0.0001pt 0.5in; font-size: 11pt; font-family: =
Calibri, sans-serif;" class=3D"">of the draft.&nbsp;&nbsp; The decision =
on whether or not to proceed will<o:p class=3D""></o:p></div><div =
style=3D"margin: 0in 0in 0.0001pt 0.5in; font-size: 11pt; font-family: =
Calibri, sans-serif;" class=3D"">be based on rough consensus of the WG, =
see RFC 7282.<br class=3D"">During the 2nd WGLC, Eric Rescorla and David =
Schinazi expressed<br class=3D"">strong views that this draft should not =
be published =E2=80=93 &nbsp;those<br class=3D"">concerns have not been =
resolved and are carried forward to<o:p class=3D""></o:p></div><div =
style=3D"margin: 0in 0in 0.0001pt 0.5in; font-size: 11pt; font-family: =
Calibri, sans-serif;" class=3D"">this WGLC.&nbsp; This email message was =
an attempt to summarize<o:p class=3D""></o:p></div><div style=3D"margin: =
0in 0in 0.0001pt 0.5in; font-size: 11pt; font-family: Calibri, =
sans-serif;" class=3D"">those concerns:<o:p class=3D""></o:p></div><div =
style=3D"margin: 0in 0in 0.0001pt; font-size: 11pt; font-family: =
Calibri, sans-serif;" class=3D""><a =
href=3D"https://mailarchive.ietf.org/arch/msg/tsvwg/i4qyY1HRqKwm0Jme9UtEb6=
DyhXU/" style=3D"color: rgb(149, 79, 114); text-decoration: underline;" =
class=3D"">https://mailarchive.ietf.org/arch/msg/tsvwg/i4qyY1HRqKwm0Jme9Ut=
Eb6DyhXU/</a><o:p class=3D""></o:p></div><div style=3D"margin: 0in 0in =
0.0001pt 0.5in; font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D"">Further explanation from both Eric Rescorla and David =
Schinazi<o:p class=3D""></o:p></div><div style=3D"margin: 0in 0in =
0.0001pt 0.5in; font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D"">is welcome and encouraged to ensure that their concerns =
are<o:p class=3D""></o:p></div><div style=3D"margin: 0in 0in 0.0001pt =
0.5in; font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D"">clearly understood.<o:p class=3D""></o:p></div><div =
style=3D"margin: 0in 0in 0.0001pt; font-size: 11pt; font-family: =
Calibri, sans-serif;" class=3D""><o:p class=3D"">&nbsp;</o:p></div><ol =
start=3D"2" type=3D"1" style=3D"margin-bottom: 0in; margin-top: 0in;" =
class=3D""><li class=3D"MsoPlainText" style=3D"margin: 0in 0in 0.0001pt; =
font-size: 11pt; font-family: Calibri, sans-serif;">Review of changes =
made since the -12 version of the draft that<br class=3D"">was the =
subject of the second WGLC (e.g., whether or not they<br =
class=3D"">suffice to resolve concerns raised during that WGLC, other<br =
class=3D"">than overall objections to publishing this draft as an =
RFC):<o:p class=3D""></o:p></li></ol><div style=3D"margin: 0in 0in =
0.0001pt; font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D""><a =
href=3D"https://www.ietf.org/rfcdiff?url1=3Ddraft-ietf-tsvwg-transport-enc=
rypt-12&amp;url2=3Ddraft-ietf-tsvwg-transport-encrypt-15" style=3D"color: =
rgb(149, 79, 114); text-decoration: underline;" =
class=3D"">https://www.ietf.org/rfcdiff?url1=3Ddraft-ietf-tsvwg-transport-=
encrypt-12&amp;url2=3Ddraft-ietf-tsvwg-transport-encrypt-15</a><o:p =
class=3D""></o:p></div><div style=3D"margin: 0in 0in 0.0001pt; =
font-size: 11pt; font-family: Calibri, sans-serif;" class=3D""><o:p =
class=3D"">&nbsp;</o:p></div><div style=3D"margin: 0in 0in 0.0001pt; =
font-size: 11pt; font-family: Calibri, sans-serif;" class=3D"">Comments =
should be sent to the<span class=3D"Apple-converted-space">&nbsp;</span><a=
 href=3D"mailto:tsvwg@ietf.org" style=3D"color: rgb(149, 79, 114); =
text-decoration: underline;" class=3D"">tsvwg@ietf.org</a><span =
class=3D"Apple-converted-space">&nbsp;</span>list, although purely<o:p =
class=3D""></o:p></div><div style=3D"margin: 0in 0in 0.0001pt; =
font-size: 11pt; font-family: Calibri, sans-serif;" class=3D"">editorial =
comments may be sent directly to the authors. &nbsp;Please cc: the<o:p =
class=3D""></o:p></div><div style=3D"margin: 0in 0in 0.0001pt; =
font-size: 11pt; font-family: Calibri, sans-serif;" class=3D"">WG chairs =
at<span class=3D"Apple-converted-space">&nbsp;</span><a =
href=3D"mailto:tsvwg-chairs@ietf.org" style=3D"color: rgb(149, 79, 114); =
text-decoration: underline;" class=3D"">tsvwg-chairs@ietf.org</a>&nbsp; =
if you would like the chairs to<o:p class=3D""></o:p></div><div =
style=3D"margin: 0in 0in 0.0001pt; font-size: 11pt; font-family: =
Calibri, sans-serif;" class=3D"">track such editorial comments as part =
of the WGLC process.<o:p class=3D""></o:p></div><div style=3D"margin: =
0in 0in 0.0001pt; font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D""><o:p class=3D"">&nbsp;</o:p></div><div style=3D"margin: 0in =
0in 0.0001pt; font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D"">No IPR disclosures have been submitted directly on this =
draft.<o:p class=3D""></o:p></div><div style=3D"margin: 0in 0in =
0.0001pt; font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D""><o:p class=3D"">&nbsp;</o:p></div><div style=3D"margin: 0in =
0in 0.0001pt; font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D"">Thanks,<o:p class=3D""></o:p></div><div style=3D"margin: 0in =
0in 0.0001pt; font-size: 11pt; font-family: Calibri, sans-serif;" =
class=3D"">David and Wes (TSVWG Co-Chairs =E2=80=93 Gorry is recused as =
a draft author)</div></div></div></blockquote></div><br =
class=3D""></body></html>=

--Apple-Mail=_07F27AFB-2557-4BD3-A8FF-4AE38A9E3021--

--Apple-Mail=_0047D1F0-699F-428F-932A-6947D02C370C
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
	filename=signature.asc
Content-Type: application/pgp-signature;
	name=signature.asc
Content-Description: Message signed with OpenPGP

-----BEGIN PGP SIGNATURE-----
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=KYY6
-----END PGP SIGNATURE-----

--Apple-Mail=_0047D1F0-699F-428F-932A-6947D02C370C--


From nobody Fri Jun 12 04:48:53 2020
Return-Path: <julian.reschke@gmx.de>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id B10063A0F7A for <quic@ietfa.amsl.com>; Fri, 12 Jun 2020 04:48:52 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, FREEMAIL_FROM=0.001, RCVD_IN_MSPIKE_H2=-0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=gmx.net
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id I4MZsqNv2XNT for <quic@ietfa.amsl.com>; Fri, 12 Jun 2020 04:48:51 -0700 (PDT)
Received: from mout.gmx.net (mout.gmx.net [212.227.17.20]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id E46F53A0F79 for <quic@ietf.org>; Fri, 12 Jun 2020 04:48:50 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=gmx.net; s=badeba3b8450; t=1591962526; bh=n4RFIuhsfa1sgKxNbz3Hrtd8n0hcDRGkvFV9+1P2vIo=; h=X-UI-Sender-Class:To:Subject:From:Date; b=Jnv3HkVtfDyt1dZ0wlA0epIYAJgh6w7qHv0XINad+7piBbRRsrSiB5MJdSmISuZMP WbJDNVDHA2ZcytDSINpZuQ+EJr8pDnh/CMvNpmVEsM6mt+k59gcH7ut5pxtRoPA/V6 z3R+NmC8wt6DhBuAmJGllsrzwISFhCDMexoeotqw=
X-UI-Sender-Class: 01bb95c1-4bf8-414a-932a-4f6e2808ef9c
Received: from [192.168.178.182] ([84.171.150.215]) by mail.gmx.com (mrgmx105 [212.227.17.168]) with ESMTPSA (Nemesis) id 1MatRT-1jC0ur30hv-00cS4a for <quic@ietf.org>; Fri, 12 Jun 2020 13:48:46 +0200
To: quic@ietf.org
Subject: draft-ietf-quic-http-29, "1.1. Prior versions of HTTP"
From: Julian Reschke <julian.reschke@gmx.de>
Message-ID: <8c747b59-1672-4ccb-dfce-0914e74c504f@gmx.de>
Date: Fri, 12 Jun 2020 13:48:42 +0200
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:68.0) Gecko/20100101 Thunderbird/68.9.0
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: quoted-printable
X-Provags-ID: V03:K1:ec2u1FpcRFuSHsFszE9GRfSxWx8NBfN7Mk1zUHq2SIWPL7s0CB5 g71wN0qdG5UgfSmgTPiE+aPLaujBR58GpJvsOVciAiA9qSlyKLypIO7L39DtgRFGFkD/DAd 0bGxlwxnagcOk/NoWqW4mQPNV4PYJmlSOUelLx3eMwqob2ViTRIrgy1JbyWv0kLfadLwkG7 7jFJbDFhirdNog6oFGVUQ==
X-UI-Out-Filterresults: notjunk:1;V03:K0:jay4BgKNMT8=:FDK5Yahr5lOMjUq77osyQJ Qbh2WXOuLz6z1akkLYpjMBzrUTdaMBRwpTJZDWys68wkGVybX17tHmndOfzPGagXu0eoYlQjn wrhGJg1WNovz3KS8A6YLGBD6ujIj6DsAFpVDXOOpPy1GLag1burblvpNhEJPbc3uuyOwlgBlR C/zmKS+EORz3pI+Abyoyp65MseTZJetc3xe0PUel1heQ3JGWHBk0j0bfIHEMherZ3uSlr6nYJ Et9D5proBs93zEqSGPvjcrTZqjJiHC0KkJf3VoE5TgtRYEeuYbCoogcSwrJXSWNVo0STp+7/y u2nWrplwrdgNiFL3p0PaP071gFDF6j+TaDZ5MQzzSyUmE9srY6j3emfPEmrwH7d3Mh7/92th7 Oyq7kyv5yKBtxc3H2k5deGoW+QqBGoHM+cYmAhtSU4S1ZHBFa8+v6+mSdA60dRfqPB4tS5GN9 xnwVQBXu0tTiszD16mEctmmDb0THUETSeiNqzHzjA5Q2TIefKs731i8dUKguEP92UiPyTPy04 y0Fnpt8lmshoHPRIs/jHLunTJ+EZMktVfX6U9UL+0djmGLub3AV9Po0dAA/MvM5cadq+NuoIa Ecf0K+8TFfftkv4g6Pct6k5HDugOgiCoycYJ3LQZgdKRT/bLqtMJyMzk/xwGViX1L2oyDMxs+ uzy/kfgR1o+X9NLpZVzLT67viZlGdj2HjYB8UvyEbbvdIdTZxUH6Y27jWcqDnyJwcJRL2X6EJ K/IXcTgB8NsfY8Y5+GhPAgCqY63uIn6R74+ZzZ/5BebmFEVatRkcUx+e3uG4zeeVct8vWWbRI 4OHxldq9rTK3sNrniNzcNpKvpDVuJcSLAdty1hU8wgGAhfZyA8fRXzz7oo3sCuVmqYGv+ZFKc +JWR6F+MIySsucDfLHLxoIMcAuWBmkEc4W8IXJtcLxHhXIlhc/qn9nnM4vapmN1Gpj4tbw29E n0SfMOSyVzPT5cleT3AbNKhW/ybpNjNqTM+rKcwJSZ8YrWADRbGXO3to82qPuZCcVP9M/YWfq v+uqvaNirWwpQN2uEHEsm2VE8kyQ35KKqJLWib+NKp1ktX2i0nMUZkOiq1ca3VrDP5wCof6aZ curvJDHvVJPQv3lOFvWcQ9WBCvYxr8MtLq8z8TpNqcLVNgtsHc5+OvDx6DdTPelfc0LxgNrJe NaM1K58S3CQRed8SCy5lf6CNVEr/bNKZBXEAYsT25TZNoYtwW2rbukeYHmZP/RYWr7rboSgqS d+IW6+OWz1nhT9AhL
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/3sXpTSYdT5NXv9N48DsMxOlvoRI>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 12 Jun 2020 11:48:53 -0000

> HTTP/1.1 [HTTP11] is a TCP mapping which uses whitespace-delimited text =
fields to convey HTTP messages. While these exchanges are human-readable, =
using whitespace for message formatting leads to parsing complexity and mo=
tivates tolerance of variant behavior. Because each connection can transfe=
r only a single HTTP request or response at a time in each direction, mult=
iple parallel TCP connections are often used, reducing the ability of the =
congestion controller to effectively manage traffic between endpoints.

Isn't that a bit misleading, as at least in theory HTTP/1.1 supports
pipelining?

BEst regards, Julian


From nobody Fri Jun 12 05:36:21 2020
Return-Path: <julian.reschke@gmx.de>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id D4D4C3A0FAD for <quic@ietfa.amsl.com>; Fri, 12 Jun 2020 05:36:16 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.899
X-Spam-Level: 
X-Spam-Status: No, score=-1.899 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, FREEMAIL_FROM=0.001, RCVD_IN_MSPIKE_H2=-0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=gmx.net
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id HEnaxXh67xqF for <quic@ietfa.amsl.com>; Fri, 12 Jun 2020 05:36:15 -0700 (PDT)
Received: from mout.gmx.net (mout.gmx.net [212.227.17.20]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id E2B273A09A0 for <quic@ietf.org>; Fri, 12 Jun 2020 05:36:14 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=gmx.net; s=badeba3b8450; t=1591965371; bh=/aMf45qKHNcVY3NVQiAYZC1E+Z6P37H4J+dRC1NUlhE=; h=X-UI-Sender-Class:To:Subject:From:Date; b=i6GPAfSnNNjRA3VkD3xXel/nwABWzCrv0av3bJa97vUwwo0v9v58s0dG4dMxSFzvN FCX8G+UhoIx7bm8IKB/CB93Md5By6dECR9PwPZlbrZ1M4My6iTOwWG+theiW7NnV6p nf4K825XsVz7fjEdeBRrQMR2M+JTx5VSY9pckqYE=
X-UI-Sender-Class: 01bb95c1-4bf8-414a-932a-4f6e2808ef9c
Received: from [192.168.178.182] ([84.171.150.215]) by mail.gmx.com (mrgmx104 [212.227.17.168]) with ESMTPSA (Nemesis) id 1Mzyuc-1ixvBM3lsU-00x1wZ for <quic@ietf.org>; Fri, 12 Jun 2020 14:36:10 +0200
To: quic@ietf.org
Subject: draft-ietf-quic-http-29, "4.1.1. Field Formatting and Compression"
From: Julian Reschke <julian.reschke@gmx.de>
Message-ID: <243a18d0-c4b6-d19d-b0af-85daf0a4a3b4@gmx.de>
Date: Fri, 12 Jun 2020 14:36:08 +0200
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:68.0) Gecko/20100101 Thunderbird/68.9.0
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: quoted-printable
X-Provags-ID: V03:K1:9uAsvwpPEb2+8gaGXlt++3POyl6+W1wDi5FcZEr7kfItjzrZKZC 0Sd8d5D9wcxleViZqr7+K4My9CRQg6czuVZO4IWUqvLPBomsF50/5rLU83NSMP3Ggl+VrGb i4f11bmodqFId7/M2srTbZA7D5AgqG/N8sDhfDpeQoRNVKesUK5Y+iU1BqWdiWb8zTK830/ jeCxJ5majEwcIdcpyn6mg==
X-UI-Out-Filterresults: notjunk:1;V03:K0:EWRBVDSoLcc=:SvCL1W77HhLRlNKqR33XiQ AQam+rSqHrWhwrTlAuK4L3G/YRX8iK79QgehWZ+9T6GFDa6N+2ig4gSAQCbBhSLsQ4Qu6JoJa pwG/cfRTDDxifqLssc1LNaEplI96uY2caJbCIEo1RW4EzFoZod+h2egjc7F97pHsqym34FvCO Hs95n4AykPvubdwiiLwBhOn7gS6dyJSI+l7no6nDWdRD7Ga/cBB8TS+IquKgAbaUNXYQYLkOW MGnRbvAm3YPbR1Mn4PpcqHFKEYBlVytXWl5KpcjwZWkK9M+1BkOwZjwoEvLNRLuaUBAv/BFPT KlHCwVUTF4V2nr8Sg3ucae3F9wyrxG6KpprHplFAOs7od0PYfVHEeEomfqb+erwHJwLRY1pxI 18O66g5st4eJ7FpN6zdhizDQ5YYjJIUT8BoH0MCb84mcRi+Xqyknxdt1y5Hw1kHatV4sNh28o wZEbXroIg373z+Z5KdcWPFXX19mQjoPEeqrLP6dDDz8yfDYQ2jLVqZHc18AaSMItHmXsnPFqW iQDpSH302ORppYhKpDg4n9QxPaXp9ALfxuW/khJKNoFAYO7bOMGIlu1heqhQeHSNQjzWd4mJV s3FEUSWHCb44BSYUo8QyCLuW9DUaF2AZOFefAZN19IKyBNPqGmesbYbS50Ip+49KAg3jhcOrZ 9KAaztPCYgMv3tsfPJUvLhrNfGFDy9YEWDlMWc5lQ2GgizoI3re0cKd5lwpzc6tIWRWu/apt0 lLwx8N4tz1sLCLPMggV9k775bcxZaWAJl6nhIFvy7JeKsvf3uErWDOi4bHP6Qd9dTRRz/PZ94 oew/aryPJnN1nxxEQu8h0N+rUBcgmRvEQizioSm3ASN+2eme+pfA4gwme/Pl8M6OzUu5dvGKB dMUxYBOZngM1F8ZpM92KFDHfhFvQgl014HzNWy1/kItTwjQJv8Cudl0I+WarsaKCiKvSNGHKI W+MNRk+ksb20JvRMgRaxT5E1yo0xkS06Exi4wq4y47kDKDIzjsETUtdZXfm1Q1KxP+UeTnqXu veTb23Ihx/BnnP9TPvLzwgZRNe/ta0/aldcW0ZLBo6iWTiTnFcudWLxzhVYoDjL6dpDmsIdYH H6yNoGqhOgtuLE4d+NGKhUJznL9hLuDGULRqTFZ3a1DyZesAhagqbYAv2ibSogupCILaraDMT s1w2NTpeeKcqa6Q6M5/z6NTTz18Z0bGZU+hJY6cw3o63fNvB83y1GRr4tyS8TjzRHrOoC+Q/j hivjpY0fHl7Vlt8/O
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/6d424ookr7cbNsux40jDObPtkIc>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 12 Jun 2020 12:36:20 -0000

> HTTP messages carry metadata as a series of key-value pairs, called HTTP=
 fields. For a listing of registered HTTP fields, see the =E2=80=9CHyperte=
xt Transfer Protocol (HTTP) Field Name Registry=E2=80=9D maintained at <ht=
tps://www.iana.org/assignments/http-fields/>.

Maybe a comment should be added explaining that that URL will only start
working once the IANA actions for draft-ietf-httpbis-semantics have been
executed.

Best regards, Julian


From nobody Fri Jun 12 05:39:03 2020
Return-Path: <julian.reschke@gmx.de>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 6002F3A0991 for <quic@ietfa.amsl.com>; Fri, 12 Jun 2020 05:39:01 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, FREEMAIL_FROM=0.001, RCVD_IN_MSPIKE_H2=-0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=gmx.net
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id k3E-QcgH6VoK for <quic@ietfa.amsl.com>; Fri, 12 Jun 2020 05:39:00 -0700 (PDT)
Received: from mout.gmx.net (mout.gmx.net [212.227.17.22]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id DF4B33A098B for <quic@ietf.org>; Fri, 12 Jun 2020 05:38:59 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=gmx.net; s=badeba3b8450; t=1591965536; bh=yeTjayjd78IdWs361NkSgwkT1VZWELpXAUcWnI0XhGg=; h=X-UI-Sender-Class:To:Subject:From:Date; b=FCbc8o0L3TIzPJvd2MxmrNZlWkS32I/9T6lbnpY6RMBAbI3o9AP2TRHgZqYUsgFi2 X7xBu7ceAbzrxeWfO0aPpmBoAPErTtmXisngzv81Z4yMz6x8GKmUt4O8eeUBlAh4dM pjQNfC91l7zSwxaWk3ZxvlPF/QIXfr9k9/x2kkZY=
X-UI-Sender-Class: 01bb95c1-4bf8-414a-932a-4f6e2808ef9c
Received: from [192.168.178.182] ([84.171.150.215]) by mail.gmx.com (mrgmx105 [212.227.17.168]) with ESMTPSA (Nemesis) id 1Mw9UK-1iu5mG0yFo-00s3mS for <quic@ietf.org>; Fri, 12 Jun 2020 14:38:56 +0200
To: quic@ietf.org
Subject: draft-ietf-quic-http-29, "4.1.1.1. Pseudo-Header Fields"
From: Julian Reschke <julian.reschke@gmx.de>
Message-ID: <45488d72-73d8-9b24-5dd9-e50e2e055e0a@gmx.de>
Date: Fri, 12 Jun 2020 14:38:54 +0200
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:68.0) Gecko/20100101 Thunderbird/68.9.0
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: quoted-printable
X-Provags-ID: V03:K1:wpJPdHLvcRMGSKYYghrmqe3CyggG6M1p9G+GYratuaD5WLUfw7K B2BO4ISYci5AWKbt+Du/kXGWRavxkdM4KpWpexaORAaPuw/Dk0q8v5WqR5tIeXxPqDy72mc i2Y8vTfydorQhnTTOqJNqUbeitU180mSsrdw8yNxja9kg7FdOsKrR8reDAjbWIYfaFUryDy YxvmDb/76/1FoeNRlZwxg==
X-UI-Out-Filterresults: notjunk:1;V03:K0:VgZvpDwS5Gg=:B2DCk9SEh8G76GYytgX0fD NasVCO9HmX/SzXrZS542B3qOJFG+Opk1U9kMJ2f2yq2tts4JSlo8ZFHs3nbhhHw+Apcnv6Dzj kwdnv5TkTeRBxGiafhNbBp7TE4TGypR8ddJwnSfO49ISOTUKKpVLNtf+jrQCx4XuE2e8i7f1+ OzOowRssPKBLZl64ScMD73ANR3Iej+AO81GIxCIzemAlUocTDhOIFJKghUkbJiJ0f12JFN35A LYEXnB0pWZnPeMTZK6d+RCvPpcPnUQ7rGPAHUAHEth1PesGj7g8UQUMbp/cAH2tl5vBOdTA+H /ipBTJo+VUPyFIfAqG3Ojvzap3eMmvlCBxbUaxQemsH8ytBGH+tc3Rg4lOVwqdTht7f0F+D3V 1mTDEMT1gpWFGXIyXtrF/3cUElz+RNADi8QDQniDA4q1igXC9he89BslBPEU6juHHzLQ6c2r7 XMPM34buGQY5dUpG3j+L4hXw7saVtcyRHNUfs82KwAtWgneJY5IJo0X9Dv2yGbrKkwc6Vz/gU oW/JZ4S0XUqfF7YC32Td/vZfQ3KGVhs4oMq+kQ2SZkV3A6MZhRa1WuhSx92Bhau4mSWlrdnBu nY8EEUCNbAypoOUvj/7GrroJRY6sOdwDu59B4p4uLhulCku1ctbziW2FY8s9vnE5PYTPRa4cS z9t6Zie1pQVty7DWnT+R91Q9IIoMi02zqWK21xBeEOtKKWHTc6vZAhVJ47uK3k0YDVgw/yPVn JGQ/Frzh/2HM135nlPKhkjXYD+Uw2Dn2lkpQcuFlkY1CJDoC7OiUPUfZxDxvWI7023km8xFfW v2HTaUKvYHJ9rGd4kRouFomRbv1RaUFF6nHKrKBpek4LcdyQefOnXSeKlHGdKcdX+I7MawHJl H10H/tvfpexTZOwbg4Vg1XpJHY4qUqZl5qd61acNauI7j0E+CgKVFArMdnS/LRo88Hz48+Ccs ur9NN2SMiXyOoUpsVd5Bt1DNNISUBAHCtQzSrPjMwp8WwgRJ79PF1qoh7rFOzsIjh6gBPV0zQ YGi3rtqHSkXWkG9klSiYwfNwh0fqLAm5rzjrL5lpnWXgLt5DXioHRmPWKHCaFF4pP4otcRAoy KcUe0eJ3w8u3+xwoDXtcJfhfb079COzTyu+gHBGFvcNW3lbWO/kCXgNKxY+SD6rvTe8soLZ86 KQg0xBClTYoYM9Ytha595tH55x8fyHEbJTW5Bntu3GQMnGtoprwStnqsBkjLqx8UAtsebSSNf w9YG0O1loo6+RlHJs
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/BjMWFa7Zm7nT9AcnV7Wz_F2rMNE>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 12 Jun 2020 12:39:01 -0000

 > Like HTTP/2, HTTP/3 employs a series of pseudo-header fields where
the field name begins with the =E2=80=98:=E2=80=99 character (ASCII 0x3a).=
 These
pseudo-header fields convey the target URI, the method of the request,
and the status code for the response.

Maybe it should be pointed out that these are indeed exactly the same
fields as in HTTP/2?

Best regards, Julian


From nobody Fri Jun 12 05:55:22 2020
Return-Path: <julian.reschke@gmx.de>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 5DEFF3A07CB for <quic@ietfa.amsl.com>; Fri, 12 Jun 2020 05:55:20 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, FREEMAIL_FROM=0.001, RCVD_IN_MSPIKE_H2=-0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=gmx.net
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id qvpJnmK_ao0E for <quic@ietfa.amsl.com>; Fri, 12 Jun 2020 05:55:19 -0700 (PDT)
Received: from mout.gmx.net (mout.gmx.net [212.227.17.22]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id BB4FD3A07B9 for <quic@ietf.org>; Fri, 12 Jun 2020 05:55:18 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=gmx.net; s=badeba3b8450; t=1591966517; bh=QIhydCc/H8WK3GHvBY8oucZhpnMo4nV5A5NoLxsW9Xs=; h=X-UI-Sender-Class:To:Subject:From:Date; b=ciOaeIocLcogsTu6Uxz+rjUi+BjZ7o+n/hjgY8aIdxuurqVzZofTfec2DqI/nlcrZ IFagsZw6huW1ZuuUsQ4IByKe5VUXuQhIiJqag83C92y9GZatHC+iFvWjojUxfa6+zU fiKN5ZKqhijqCBbSY8EV5dR3er/L00mBZHpxB5EA=
X-UI-Sender-Class: 01bb95c1-4bf8-414a-932a-4f6e2808ef9c
Received: from [192.168.178.182] ([84.171.150.215]) by mail.gmx.com (mrgmx105 [212.227.17.168]) with ESMTPSA (Nemesis) id 1MmDIu-1j1po50CpM-00iGbO for <quic@ietf.org>; Fri, 12 Jun 2020 14:55:17 +0200
To: quic@ietf.org
Subject: draft-ietf-quic-http-29, "4.1.1.2. Field Compression"
From: Julian Reschke <julian.reschke@gmx.de>
Message-ID: <4e549e98-a008-749d-f2bc-06c3979fbd19@gmx.de>
Date: Fri, 12 Jun 2020 14:55:15 +0200
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:68.0) Gecko/20100101 Thunderbird/68.9.0
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: quoted-printable
X-Provags-ID: V03:K1:xwWu9kjN+nfp40jbLCivy6u1AWSjP5ggOrPR7Sa8XD9DN5OFeXo eOiftwND/wmm5Rs7hTzb+zTL90JWo2RRAXcWjWDWq8go3IlCJss1hlPPDP2AyYruoHhyaF7 QFxkObfP0BRbdLygdygOxffyiAursqCAJmZAevmX3t4UCHzZpVgdjWV/evR7gNpU/6osCzO CEAJsTR+Of3e7sx7izZow==
X-UI-Out-Filterresults: notjunk:1;V03:K0:35LqkAh26No=:Qx71jA/R1w2O5TI3a+U/r3 jT60nJiSMs8NNC4/r/n/kS6d/uaWwltSfYpwuA3F3MUVTbvEJGIDbjLORXq32DFc6VLVlz856 PfJ/AbcxeUcXRhsxEPneCV3Ip6gBpt7HLnnH9VhTIdcqDdCwaxzkRbf5y3pHFjGuojS4tEm7d mdwjcs9z1ONX2vOWtWWc4X02SVZF1h2i5dTVj+qQq1Ll/slgJBMwr4iykbqnkCKYahZc+2+JC flOUR3k4Ky3uc3taGClcegmFuzlvPS4dL+8uaviUrouRE82GOvuAXwoL2sUSZGQQWoO1YzB6S kYi6uD7DbJVwLEz9kaNpDDSE7YhVjfhpXCdtRrmfefW5m/8DOis3YYCLCZ3j+PKrs5xdf/khQ DDQDyQ819VrJrjsRZGjicliKd1sugSF1kptQWDuwoa7PEMinh0HECtWpVvXLaNjPPueBRW048 lR7vXe8ZuUO7z+tej2Pw9KfiEQyhGVO4wn97GCKY33ynuJ432FnaUkU2Lnw1HlejVHIiD6ofy 022gvs00W/ddXBYIvSPcruN3lDsb1ZbWPQ+/wE0BrR3lVMUsI+mRpDCxFx6b7AJhZqnpoFDrT ayD5mIT/V4Q6+t5o2dpnFvViIzSyVc9+iECi/R9FHUxa9xSRUqu3fywHzWlmkOPZaPndQxYHY G5HDnC5u7iqYe9EAQUPrfCbgDtIF76o/knJP4MUJPpnzod8V0pP1y9e+GF1ae61Gra3XqLgr4 ahl8+PqX8b7meWNjC6q7YJi1lQg2PgYTtexoPxwJImYbKlqXTfv/sO6G33YAga8wDTk0nXSDM zg2UCApA4DQVSSpOSKZSN+SS0RMdr6sFXKWfHFn/K5xFv9jhPvI+ijsz/12crDF74tgUSYd9o NLc45GIB1e8wr/HDPu3Lqka3YNBru7Yvi1p2kBCFZjTgv6XrfwSe0i+KNlK92/DxZ9tRbYNFR seOixmWY79DVPE2jshpI34yvJWTSUYVPtUyL4GqVzdb7e0vCsbwra2HjzJKVrt6bSQjUqr8l3 tPjqIDa5hRqXw9+UWViMohV2zWlv1xP2RRkZWBrWr1gil7KGdOiJ4vJ0DXM+PpkJb4Q9bT4+g oRTet3aNEipceeYJNnFFLHqHIsg7RC6/rP3mqoYIeWQAPF5AAMTyMAWJ0DbCCAed51S6sqKRy BZCAP9GDqPze00ZVlD8yhdIg/t82FtiqgZHwcZLdrN3qpomI31AMFDzMVxScE75IvVaTSmwVV gHeUftPPaiR3xni33
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/nqjPy0VmMY6ApXvzbXYJlxDNvY8>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 12 Jun 2020 12:55:20 -0000

 > To allow for better compression efficiency, the =E2=80=9CCookie=E2=80=
=9D field
[RFC6265] MAY be split into separate field lines, each with one or more
cookie-pairs, before compression. If a decompressed field section
contains multiple cookie field lines, these MUST be concatenated into a
single octet string using the two-octet delimiter of 0x3B, 0x20 (the
ASCII string =E2=80=9C; =E2=80=9C) before being passed into a context othe=
r than HTTP/2
or HTTP/3, such as an HTTP/1.1 connection, or a generic HTTP server
application.

Maybe it would be good to mention Section 8.1.2.5 of RFC 7540, which
defines the same for HTTP/2, but adds motivation and example.

Best regards, Julian


From nobody Fri Jun 12 06:20:36 2020
Return-Path: <robin.marx@uhasselt.be>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 61A353A090F for <quic@ietfa.amsl.com>; Fri, 12 Jun 2020 06:20:35 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, HTML_IMAGE_ONLY_32=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=uhasselt.be
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id lj1UB1dGVEkc for <quic@ietfa.amsl.com>; Fri, 12 Jun 2020 06:20:33 -0700 (PDT)
Received: from mail-wr1-x42c.google.com (mail-wr1-x42c.google.com [IPv6:2a00:1450:4864:20::42c]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id EB4093A08D4 for <quic@ietf.org>; Fri, 12 Jun 2020 06:20:32 -0700 (PDT)
Received: by mail-wr1-x42c.google.com with SMTP id e1so9730884wrt.5 for <quic@ietf.org>; Fri, 12 Jun 2020 06:20:32 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=uhasselt.be; s=google;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=SIrkYKEutV5rbE/CJZpQSDbBO3zjwi3c/Qf9bROdXCo=; b=qdV/7rpMakFSLuTk4QtaOZowY/UH+2OuE/9zCmXMsxpfm+TiEG9sSreQ0pPTk0PAON cw85OOmiw7Y7vSKnOT4n4xAqFdluF/CCG9zgxq6UfqG1x2hlGPnKhhjO+o7z4Bz7L7SO IK0uzjNUNc7QorApIykSraRG+92Sq8uqaDtOs=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=SIrkYKEutV5rbE/CJZpQSDbBO3zjwi3c/Qf9bROdXCo=; b=cRKvpZgr1T4ulfkWBQHw8YIqe778ClzJjniK7LnUOsQ7xwbCbyzHK865xJ27cbW4by s+CHcS00Th2GE3lE8PNK9FUYVXi4kpUayEfpV+NhLRbtHcW2xiDVsqFAzjNuX+VjmNJI fb8jvcxvpWbfT1L7PpQJZfl3TrvxbAHNBP54Sv/bchplvWOjPt8HkYOw730ap7N68dr4 EyNMqiucpNhFOwvpMqKfNUnDLz15ng8MfFBpR9Twr2qeeav0lT2KWTlnUU3jfqW8keQS K3NebJmvlFGz3BfMdzVPhaE87ueXh5GFTAbjdbnROg0x+RzMA8cmX0eTFBp7plnf2RPA TCdw==
X-Gm-Message-State: AOAM530D/liZ3PhbuRInUw+oXshvmyzbOk97u8H3o0Eprq+j/dNPY8Vz MbHV6D0PJaRXNkavc1t7ZPo3HIEG5uIqJQ3LkF1C9lmI
X-Google-Smtp-Source: ABdhPJxI3VllcuhtlTWF+WkA5WRFJUJyHHbheYSadfSSJDld4cJa6k2LkrPvcWwsZ6yDoZgazgYhGWuvXQvMmjBMyRU=
X-Received: by 2002:a5d:4dd0:: with SMTP id f16mr15902225wru.117.1591968031313;  Fri, 12 Jun 2020 06:20:31 -0700 (PDT)
MIME-Version: 1.0
References: <8c747b59-1672-4ccb-dfce-0914e74c504f@gmx.de>
In-Reply-To: <8c747b59-1672-4ccb-dfce-0914e74c504f@gmx.de>
From: Robin MARX <robin.marx@uhasselt.be>
Date: Fri, 12 Jun 2020 15:20:19 +0200
Message-ID: <CAC7UV9bmX2Ta5dgazge1NQdpCW2UshVvy4fUD4pwF+=BNLW5Qw@mail.gmail.com>
Subject: Re: draft-ietf-quic-http-29, "1.1. Prior versions of HTTP"
To: Julian Reschke <julian.reschke@gmx.de>
Cc: IETF QUIC WG <quic@ietf.org>
Content-Type: multipart/alternative; boundary="000000000000002cdc05a7e2f133"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/x7inn0kwWzmBumeJTepYl67WoyY>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 12 Jun 2020 13:20:35 -0000

--000000000000002cdc05a7e2f133
Content-Type: text/plain; charset="UTF-8"

Hello Julian,

To me, the crux is in the "at a time", which I feel makes the statement
correct.
Of course the nuances might not be clear for less experienced readers, but
I don't feel the H3 text should explain the details of this.

With best regards,
Robin

On Fri, 12 Jun 2020 at 13:49, Julian Reschke <julian.reschke@gmx.de> wrote:

> > HTTP/1.1 [HTTP11] is a TCP mapping which uses whitespace-delimited text
> fields to convey HTTP messages. While these exchanges are human-readable,
> using whitespace for message formatting leads to parsing complexity and
> motivates tolerance of variant behavior. Because each connection can
> transfer only a single HTTP request or response at a time in each
> direction, multiple parallel TCP connections are often used, reducing the
> ability of the congestion controller to effectively manage traffic between
> endpoints.
>
> Isn't that a bit misleading, as at least in theory HTTP/1.1 supports
> pipelining?
>
> BEst regards, Julian
>
>

-- 

Robin Marx
PhD researcher - web performance
Expertise centre for Digital Media

T +32(0)11 26 84 79 - GSM +32(0)497 72 86 94

www.uhasselt.be
Universiteit Hasselt - Campus Diepenbeek
Agoralaan Gebouw D - B-3590 Diepenbeek
Kantoor EDM-2.05

--000000000000002cdc05a7e2f133
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Hello Julian,<div><br></div><div>To me, the crux is in the=
 &quot;at a time&quot;, which I feel makes the statement correct.</div><div=
>Of course the nuances might not be clear for less experienced readers, but=
 I don&#39;t feel the H3 text should explain the details of this.=C2=A0</di=
v><div><br></div><div>With best regards,</div><div>Robin</div></div><br><di=
v class=3D"gmail_quote"><div dir=3D"ltr" class=3D"gmail_attr">On Fri, 12 Ju=
n 2020 at 13:49, Julian Reschke &lt;<a href=3D"mailto:julian.reschke@gmx.de=
">julian.reschke@gmx.de</a>&gt; wrote:<br></div><blockquote class=3D"gmail_=
quote" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,=
204);padding-left:1ex">&gt; HTTP/1.1 [HTTP11] is a TCP mapping which uses w=
hitespace-delimited text fields to convey HTTP messages. While these exchan=
ges are human-readable, using whitespace for message formatting leads to pa=
rsing complexity and motivates tolerance of variant behavior. Because each =
connection can transfer only a single HTTP request or response at a time in=
 each direction, multiple parallel TCP connections are often used, reducing=
 the ability of the congestion controller to effectively manage traffic bet=
ween endpoints.<br>
<br>
Isn&#39;t that a bit misleading, as at least in theory HTTP/1.1 supports<br=
>
pipelining?<br>
<br>
BEst regards, Julian<br>
<br>
</blockquote></div><br clear=3D"all"><div><br></div>-- <br><div dir=3D"ltr"=
 class=3D"gmail_signature"><div dir=3D"ltr"><div><br><table border=3D"0" ce=
llpadding=3D"0" cellspacing=3D"0" style=3D"width:600px;vertical-align:top;f=
ont-size:12px;color:rgb(0,0,0);text-align:left;padding:0px;font-family:Verd=
ana,Geneva,sans-serif"><tbody><tr><td style=3D"font-weight:bold;font-size:1=
6px;color:rgb(231,59,43);line-height:1.1em;padding:0px;font-family:Verdana,=
Geneva,sans-serif">Robin=C2=A0Marx</td></tr><tr><td style=3D"font-family:Ve=
rdana,Geneva,sans-serif">PhD researcher - web performance</td></tr><tr><td =
style=3D"font-family:Verdana,Geneva,sans-serif">Expertise centre for Digita=
l Media</td></tr><tr><td>=C2=A0</td></tr><tr><td style=3D"font-family:Verda=
na,Geneva,sans-serif"><span style=3D"color:rgb(231,59,43);font-weight:bold;=
font-family:Verdana,Geneva,sans-serif">T</span> +32(0)11 26 84 79 - <span s=
tyle=3D"color:rgb(231,59,43);font-weight:bold">GSM</span> +32(0)497 72 86 9=
4 </td></tr><tr><td>=C2=A0</td></tr><tr><td style=3D"font-family:Verdana,Ge=
neva,sans-serif"><a href=3D"http://www.uhasselt.be" target=3D"_blank"><span=
>www.uhasselt.be</span></a></td></tr><tr><td style=3D"font-family:Verdana,G=
eneva,sans-serif">Universiteit Hasselt  -  Campus Diepenbeek</td></tr><tr><=
td style=3D"font-family:Verdana,Geneva,sans-serif">Agoralaan  Gebouw D  -  =
B-3590 Diepenbeek</td></tr><tr><td style=3D"font-family:Verdana,Geneva,sans=
-serif">Kantoor=C2=A0EDM-2.05</td></tr><tr><td>=C2=A0</td></tr><tr><td styl=
e=3D"font-family:Verdana,Geneva,sans-serif"><div style=3D"width:100%;text-a=
lign:left"><img src=3D"https://www.uhasselt.be/images/nieuwsbrief/dcm/UHass=
elt-liggend.jpg" style=3D"float: none;" alt=3D""></div></td></tr></tbody></=
table></div></div></div>

--000000000000002cdc05a7e2f133--


From nobody Fri Jun 12 07:21:42 2020
Return-Path: <julian.reschke@gmx.de>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 49E103A0820 for <quic@ietfa.amsl.com>; Fri, 12 Jun 2020 07:21:40 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, FREEMAIL_FROM=0.001, RCVD_IN_MSPIKE_H2=-0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=gmx.net
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id VxnKtjqc368T for <quic@ietfa.amsl.com>; Fri, 12 Jun 2020 07:21:38 -0700 (PDT)
Received: from mout.gmx.net (mout.gmx.net [212.227.17.22]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 306AA3A07A9 for <quic@ietf.org>; Fri, 12 Jun 2020 07:21:38 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=gmx.net; s=badeba3b8450; t=1591971695; bh=R5qjr9NWQdE1fqikesJwPoDMbnGEMrEIGXNrrTu1dIg=; h=X-UI-Sender-Class:Subject:To:Cc:References:From:Date:In-Reply-To; b=lV84/MpGcxAefOflYfdziUiOU927p+VmIGv+x5gOVEkNPODzALILsHmD21KRxGoOf 9vvYPAapzf01+ZFT0fyRgQVsqQgOxtZd+4i9eIMLtR1QkXQLcRBaiaHrWG+4w7p2PG G0UsR3oNH+155k5/r0EczfFzcwZmRpx6IhfrfEo4=
X-UI-Sender-Class: 01bb95c1-4bf8-414a-932a-4f6e2808ef9c
Received: from [192.168.178.182] ([84.171.150.215]) by mail.gmx.com (mrgmx105 [212.227.17.168]) with ESMTPSA (Nemesis) id 1Mirng-1jFvad1OkO-00ev4I; Fri, 12 Jun 2020 16:21:35 +0200
Subject: Re: draft-ietf-quic-http-29, "1.1. Prior versions of HTTP"
To: Robin MARX <robin.marx@uhasselt.be>
Cc: IETF QUIC WG <quic@ietf.org>
References: <8c747b59-1672-4ccb-dfce-0914e74c504f@gmx.de> <CAC7UV9bmX2Ta5dgazge1NQdpCW2UshVvy4fUD4pwF+=BNLW5Qw@mail.gmail.com>
From: Julian Reschke <julian.reschke@gmx.de>
Message-ID: <672c5568-e137-0105-b043-baba983a289e@gmx.de>
Date: Fri, 12 Jun 2020 16:21:33 +0200
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:68.0) Gecko/20100101 Thunderbird/68.9.0
MIME-Version: 1.0
In-Reply-To: <CAC7UV9bmX2Ta5dgazge1NQdpCW2UshVvy4fUD4pwF+=BNLW5Qw@mail.gmail.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: quoted-printable
X-Provags-ID: V03:K1:Uocq1+UZbj3LkfiiZ8FerPgVfUgdso3Zd9J8WdupcUOLQDIuv5K 8cvTKWj3HH2sJOmyXLo4XgGjO5+Xsk+MJl+LYG+BGx7+93YNmYCby4HRc1QDU4pak09J3mb 8csq0+vmL5chkvr2vYXO6D9ynfJ+We/UIp+QIGvRQOaM5HWrrJ9P3QfBovgnYQxbZ2pUcxl BAfDTWM1L3leq/cr0tFSA==
X-UI-Out-Filterresults: notjunk:1;V03:K0:Xw3bpvOipso=:ZDXKXWmaYwPNPOpe2a/p0Z 6hrfDFnPbEztGuFP48zc6uuhUELox+zAg8BQJjCvQ3nsSYXRpCXwMQFicogGaU/wvmU7+9nS7 h66/11J5ZQMVY3nN8DKqq2JUHKmVtBRwOkOzM2lL66/onEWAfGr3VREALgutaa8D8V9ZHa71a X2AghcZj3WRYaHNbb5+w/7g4ndCxII2y9dSwu69J1DqSlr3kHcEcPefWERI85oj2P1f8nDB5l bfKUVcGEA+OqjAddda5Yshqmlvdw5MEfg8LqdyOl8ZPYR5CyzU+4xW80llsLbVyEJBOuBzoQi gz+vbdLku708x4BeSXgb6+DfRyPisjN1G4ekgLuTMC4epTCbkWe2fg36jALmmS2pMe1JN9BMZ QNjtf7h8U3d1Mof9tmAirejcUw6jrspNxbfHLgee2RWQj6NfIPgpWnRNWfgE9CKxj/Yxm9546 pEBUs+3TakjQQOyJ/2RL/O6VG+lgkWM6uu4jig5orDotwssvk6RlnKRbBLGpCTHUtfwYcIG4D OPVH8Hm3wVykTdFDS1uEA8JN/Ag9je35yS6Zn4KKzfygE6dDodr3wzjdGmbENOvUNvAnSEWUC fuGm7q0Jkz8AX0k6GKTs7Bs+5ILaKSacFH8cEbvdkqCfHsMseVcUBO/yf/ei2kJqSea9ImkTx VlRUCkJV+LWHgU5is6muBkCvb3g606XYyPxLUHQvAbrLkL7rWh24s4RLIrvDoM+mL9JEwfxiG keJVu6YX9raQsB8J9CcmX0EGhAt6AZGkLpqRHCX/x+2bYWRPQ8/RLyGGE+bDFGqVUEqBQMKTf jmZwERHaHWx55BK2ewRyfvs1Qwu4MaCM3WjqZt6l7Mw3Dm191stOJDg0Rkcjf7C3XXRUUMR1R moBdvVGgV4Lw+huuelMRkd0jLR6s9iXe6hYjCNQnFjQU5wFcEf2B7ChBwj9XlgtWa25cqB+Ck TeNf9wSDVkPhkX6t7gw866qIOzR3u/lQ4uxtB7t9C85xwgXV48UyGOjUztFuaMvsLTRkrPKm4 S6MwLLEV7EHhehVxWcqCOMX/2Gs2nca5lr4bQ45YqK/j+mkts38O0I2dELwgSaFfl+iJ6dunN QAJYZtQ/VPCvgPn/wtGlZD8wTNSEjGgXBun+X8WewP6c3cm8k1JzuaqD9fUILnHawNRYWx2oB cFYzTDQcREb1yuw0ujMrpQgMMa18DBquhZz+4yg7QwIS46Z9Ea85xhSGYubm5rU7ktXyxyPsB ZYyZLUy3+iIQ40FFn
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/cx-p37ObrQFqYEGl_raIF-47IGc>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 12 Jun 2020 14:21:40 -0000

On 12.06.2020 15:20, Robin MARX wrote:
> Hello Julian,
>
> To me, the crux is in the "at a time", which I feel makes the statement
> correct.
 > ...

Not really; if the server is not broken I can send multiple GET requests
without having read any response yet.

Best regards, Julian


From nobody Fri Jun 12 08:55:42 2020
Return-Path: <mbishop@evequefou.be>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id E937F3A08E9 for <quic@ietfa.amsl.com>; Fri, 12 Jun 2020 08:55:39 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.901
X-Spam-Level: 
X-Spam-Status: No, score=-1.901 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, RCVD_IN_MSPIKE_H2=-0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=evequefou.onmicrosoft.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id SXqHAphyiXZq for <quic@ietfa.amsl.com>; Fri, 12 Jun 2020 08:55:38 -0700 (PDT)
Received: from NAM02-SN1-obe.outbound.protection.outlook.com (mail-eopbgr770129.outbound.protection.outlook.com [40.107.77.129]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 243373A08C9 for <quic@ietf.org>; Fri, 12 Jun 2020 08:55:37 -0700 (PDT)
ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=PrsC83vPIdISVDn/GMQ+tEZwSdx/sRVutGUrQJI9mK0TySWGaIbtp4yj+oRdDZn0d8BPg1LpqotRo5K9+DAhoICDDKwAewQi5FXQPEhV9RBkQdAVRdDHwT0crYxU96p0JRNzVfQIn/SQlAgsBCw+3XJ7l9u6Xki/uRSMjSc79vfVLDOlkSVvOXsLXJJXOjyV9p661VaLtCXwVuYFbNPFp4AGO559pyTdedlVXHxycuu8lDuzRz28y0y/d63Bk0zs+RHnZ39mTHXutoejubr+Q/gq/dMbdV/ouR12AR34z1BeEPNAVymEFJ3ZQjR1UfeC+kMorrh8xsH0/YWe7gOFQg==
ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com;  s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=Y2R5NexZdcXT2I6S7jyShHBF7R9ePc7hF3O4mqLNvtg=; b=IU3NKHN0zA3RoSBV/DgZZxZwYo7t6EN/oSdsK//fqnVlPM1XVnUezWFpAg1YpRHUm/izHrpUmb3+vYJW2qJQevkGTefnFme4QerLlS8O65TCKzNVan9XexHXV0xEtI5MAcLesGqbx/086TBumJM4QBfDKKjARguQH5mYsAis5PMhi20YPvw14OfDzd5WqN9rhfYeQkFG5KQbG/ZFChtvT8wQP0/IzzuomQ2DQzAkIo+dM+wPeLHj6CScj25cLKwA04AmkLefp2GKfSaKCoSlrS9NikC4GDfHkjq7DnbFCcFtUJNB6TiOORaorQTWEPFMt12DSulfiwgKdLWEYrqKJg==
ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=evequefou.be; dmarc=pass action=none header.from=evequefou.be; dkim=pass header.d=evequefou.be; arc=none
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=evequefou.onmicrosoft.com; s=selector2-evequefou-onmicrosoft-com; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=Y2R5NexZdcXT2I6S7jyShHBF7R9ePc7hF3O4mqLNvtg=; b=FBDUEG2tKMz066eJBz3SvGj2RWt5np2OJe6ncU76Gu7EfCtTrj2pfE+4KAg06CPuLCQ7EGiT42pASFOiuwM1WUitOYsuTE+FHnlxVSxOh3WFZsB1lEEylxAzDaahbQfPMMbeTjZx7aEwof+oLzSC9iDsbhwQpX9AXRpLiQw4jJo=
Received: from CH2PR22MB2086.namprd22.prod.outlook.com (2603:10b6:610:8c::8) by CH2PR22MB2069.namprd22.prod.outlook.com (2603:10b6:610:88::19) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.3088.22; Fri, 12 Jun 2020 15:55:36 +0000
Received: from CH2PR22MB2086.namprd22.prod.outlook.com ([fe80::f0e6:8202:6056:a596]) by CH2PR22MB2086.namprd22.prod.outlook.com ([fe80::f0e6:8202:6056:a596%3]) with mapi id 15.20.3088.021; Fri, 12 Jun 2020 15:55:36 +0000
From: Mike Bishop <mbishop@evequefou.be>
To: Julian Reschke <julian.reschke@gmx.de>, Robin MARX <robin.marx@uhasselt.be>
CC: IETF QUIC WG <quic@ietf.org>
Subject: RE: draft-ietf-quic-http-29, "1.1. Prior versions of HTTP"
Thread-Topic: draft-ietf-quic-http-29, "1.1. Prior versions of HTTP"
Thread-Index: AQHWQK9/mMafyGA6tUCJnaX0MKo996jU9xCAgAARHICAABVNcA==
Date: Fri, 12 Jun 2020 15:55:36 +0000
Message-ID: <CH2PR22MB2086EED7405119874EA54374DA810@CH2PR22MB2086.namprd22.prod.outlook.com>
References: <8c747b59-1672-4ccb-dfce-0914e74c504f@gmx.de> <CAC7UV9bmX2Ta5dgazge1NQdpCW2UshVvy4fUD4pwF+=BNLW5Qw@mail.gmail.com> <672c5568-e137-0105-b043-baba983a289e@gmx.de>
In-Reply-To: <672c5568-e137-0105-b043-baba983a289e@gmx.de>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
authentication-results: gmx.de; dkim=none (message not signed) header.d=none;gmx.de; dmarc=none action=none header.from=evequefou.be;
x-originating-ip: [2600:2b00:9309:6a01:6919:4193:fe07:f6e8]
x-ms-publictraffictype: Email
x-ms-office365-filtering-correlation-id: f5d2e3f1-5561-4b02-6d45-08d80ee90bec
x-ms-traffictypediagnostic: CH2PR22MB2069:
x-microsoft-antispam-prvs: <CH2PR22MB2069B05430A454334D2ED8FCDA810@CH2PR22MB2069.namprd22.prod.outlook.com>
x-ms-oob-tlc-oobclassifiers: OLM:2089;
x-forefront-prvs: 0432A04947
x-ms-exchange-senderadcheck: 1
x-microsoft-antispam: BCL:0;
x-microsoft-antispam-message-info: aXeZ+7y8RH7E9D68vlzxXiMrhf2pymS+8xoOkGcPj8FVmv25ZVEkUCqWBkfKSyjrjU4PkUBp41hJjhn5e6idtdCAGwIFrBjhS1cJ3HnnnXDv9DFLNaGmRlp0UwZ+4kFB+Z2xdwvnjnhK+RH46z9GWENsMiIcHBC1awVbkh0om6TOS2fKcfiDMPc6Un3o9xo4aGN9yKqy012SMncE4cEoDmY5ErFgEOYXq1nNc4l/ECdyeORIlo0eKF3dOez06JZjfYNbp+GL9fuJsdLDppoTihupzwNu1nuJRO7sLFGy1SHnm8UVV2+CAKDgaNnZJDa648Y2qiFAod6u7vsVUpD5Hg==
x-forefront-antispam-report: CIP:255.255.255.255; CTRY:; LANG:en; SCL:1; SRV:;  IPV:NLI; SFV:NSPM; H:CH2PR22MB2086.namprd22.prod.outlook.com; PTR:; CAT:NONE;  SFTY:; SFS:(39830400003)(136003)(346002)(376002)(396003)(366004)(4744005)(83380400001)(33656002)(64756008)(508600001)(76116006)(7696005)(66556008)(66946007)(66446008)(186003)(6506007)(86362001)(66476007)(53546011)(52536014)(4326008)(5660300002)(316002)(71200400001)(8676002)(8936002)(2906002)(9686003)(55016002)(110136005); DIR:OUT; SFP:1102; 
x-ms-exchange-antispam-messagedata: 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
x-ms-exchange-transport-forked: True
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
X-OriginatorOrg: evequefou.be
X-MS-Exchange-CrossTenant-Network-Message-Id: f5d2e3f1-5561-4b02-6d45-08d80ee90bec
X-MS-Exchange-CrossTenant-originalarrivaltime: 12 Jun 2020 15:55:36.2132 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Hosted
X-MS-Exchange-CrossTenant-id: 41eaf50b-882d-47eb-8c4c-0b5b76a9da8f
X-MS-Exchange-CrossTenant-mailboxtype: HOSTED
X-MS-Exchange-CrossTenant-userprincipalname: PEHYppEGGxv0jwJJsMHScfc72HBitkUg2stNLvfUT8QnX3guLUopx1IeDRszotS0M6/S9akee4wJUnFjzCPIqQ==
X-MS-Exchange-Transport-CrossTenantHeadersStamped: CH2PR22MB2069
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/3kjiXgAHZTu7oXQ6pXmm6GNVjcc>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 12 Jun 2020 15:55:40 -0000
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From nobody Fri Jun 12 11:01:30 2020
Return-Path: <paul@redbarn.org>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id AC8A63A07C4 for <quic@ietfa.amsl.com>; Fri, 12 Jun 2020 11:01:27 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, SPF_HELO_NONE=0.001, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Wyp0cMNjYy6G for <quic@ietfa.amsl.com>; Fri, 12 Jun 2020 11:01:25 -0700 (PDT)
Received: from family.redbarn.org (family.redbarn.org [24.104.150.213]) (using TLSv1.2 with cipher ADH-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 3FFC43A07BA for <quic@ietf.org>; Fri, 12 Jun 2020 11:01:14 -0700 (PDT)
Received: from linux-9daj.localnet (dhcp-166.access.rits.tisf.net [24.104.150.166]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (1024 bits) server-digest SHA256) (Client did not present a certificate) by family.redbarn.org (Postfix) with ESMTPSA id 35047B07D1; Fri, 12 Jun 2020 18:01:14 +0000 (UTC)
From: Paul Vixie <paul@redbarn.org>
To: Julian Reschke <julian.reschke@gmx.de>, Robin MARX <robin.marx@uhasselt.be>, quic@ietf.org
Cc: IETF QUIC WG <quic@ietf.org>, Mike Bishop <mbishop@evequefou.be>
Subject: Re: draft-ietf-quic-http-29, "1.1. Prior versions of HTTP"
Date: Fri, 12 Jun 2020 18:01:13 +0000
Message-ID: <1846406.p7HaY3GM0x@linux-9daj>
Organization: none
In-Reply-To: <CH2PR22MB2086EED7405119874EA54374DA810@CH2PR22MB2086.namprd22.prod.outlook.com>
References: <8c747b59-1672-4ccb-dfce-0914e74c504f@gmx.de> <672c5568-e137-0105-b043-baba983a289e@gmx.de> <CH2PR22MB2086EED7405119874EA54374DA810@CH2PR22MB2086.namprd22.prod.outlook.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 7Bit
Content-Type: text/plain; charset="us-ascii"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/yDDGj9Bmep8BbQsEc0JYlmvJEHg>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 12 Jun 2020 18:01:28 -0000

On Friday, 12 June 2020 15:55:36 UTC Mike Bishop wrote:
> Robin has almost the intended reading.  One request/response at a time, in
> each direction.  So while you might send several requests in sequence
> before getting any responses back, each direction is always either in the
> middle of exactly one message, or between messages.

the terminology here is easily confusable. in HTTP/1.1, a greedy responder can 
parse pipelined requests and start the work associated with each. however, the 
answers have to come back in order. this "head of line blocking" imposes 
response serialization, and HTTP/3 avoids this.

one way to word this is to call out serialization of both requests and 
responses, and ordering of responses, as limits of HTTP/1.1 that HTTP/3 is 
designed to overcome. wording like "at the same time" is less clear.

-- 
Paul



From nobody Sun Jun 14 00:33:00 2020
Return-Path: <do_not_reply@mnot.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 62E653A0C84 for <quic@ietfa.amsl.com>; Sun, 14 Jun 2020 00:32:58 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.197
X-Spam-Level: 
X-Spam-Status: No, score=-0.197 tagged_above=-999 required=5 tests=[DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, HTML_MESSAGE=0.001, RCVD_IN_MSPIKE_H3=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=mnot.net header.b=gfJm3r8Q; dkim=pass (2048-bit key) header.d=messagingengine.com header.b=vFkxnV4o
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id usQNs1yloIxb for <quic@ietfa.amsl.com>; Sun, 14 Jun 2020 00:32:56 -0700 (PDT)
Received: from wout2-smtp.messagingengine.com (wout2-smtp.messagingengine.com [64.147.123.25]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 05D6F3A0C83 for <quic@ietf.org>; Sun, 14 Jun 2020 00:32:55 -0700 (PDT)
Received: from compute1.internal (compute1.nyi.internal [10.202.2.41]) by mailout.west.internal (Postfix) with ESMTP id EDD3A40D for <quic@ietf.org>; Sun, 14 Jun 2020 03:32:54 -0400 (EDT)
Received: from mailfrontend2 ([10.202.2.163]) by compute1.internal (MEProxy); Sun, 14 Jun 2020 03:32:55 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=mnot.net; h= content-type:mime-version:from:to:subject:message-id:date; s= fm2; bh=mcya3JX60lUFLbDQn4RhaIDSbe1/6k4+MZJZDqySd5Y=; b=gfJm3r8Q dIOfzKpWImBZcZ9ml4PZjnYyHYKNpQ9l3YFqQB6yIfM0Fns7ql39gzK5BY9k6eIC EbL+QjlxNZFULvezfG5ak+ARqS7g539jFDRWo9ibMRw8wbEabsvWzmgKPKqg4hr0 CMxdlXgkPi5iKOn94nDBhXQiQFiCCaIC94WTmxt8LTfqTxESHdlBmrqXD+fc1o/X R4vokl/l8unRSfXof+/tl0d+P97pY030jJ2Er1aISMg4VdLKLe8WBD+GPzgzlcHp 8OYprzsBH8bRkSS6dKS/rH8cPc/0wESKp2dv9zA7k4+TGTbGDYby2HO+CwGvT+Eu REuqF0MVu42Rmg==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d= messagingengine.com; h=content-type:date:from:message-id :mime-version:subject:to:x-me-proxy:x-me-proxy:x-me-sender :x-me-sender:x-sasl-enc; s=fm3; bh=mcya3JX60lUFLbDQn4RhaIDSbe1/6 k4+MZJZDqySd5Y=; b=vFkxnV4oqa3Zlz5RaIoMjuZ5Ej0imVusvJ1pqmfWgjT3/ Y3qFJ68S5H7K64+RJMkygTDTuYWc0lNmLVRucNz1TRef7BoUAdCQEEH63aWzkbAO 2UnXs1FbeyM1tlWgU66JlZlhgjJlNh8DqjupuktXYiEnbygyKfkr9JU0+BBfyCg9 dJg2m1jBNLHB02FUnmoh+prUjjWqV8Z5qBc9DLfRCpYHTzViOXNIpse+2ba4MWae GqInGvCNyEi1POM4cj3UK5OSBOIerOAXzo9mi4Gwwz/trq28YaTfziEQlWOjvWla 5lh1TPONjJT46RyErsEDIvjfWdGy3h1kgQ+LJDhEA==
X-ME-Sender: <xms:ptLlXk6iYPZr-Ja5SOfAC93_Lf7iNXpam2lW_I6srW1sTcLXZ5Rxew>
X-ME-Proxy-Cause: gggruggvucftvghtrhhoucdtuddrgeduhedrudeihedgvdekucetufdoteggodetrfdotf fvucfrrhhofhhilhgvmecuhfgrshhtofgrihhlpdfqfgfvpdfurfetoffkrfgpnffqhgen uceurghilhhouhhtmecufedttdenucenucfjughrpegtggfhvffusegrtddtredttdejne cuhfhrohhmpeftvghpohhsihhtohhrhicutegtthhivhhithihucfuuhhmmhgrrhihuceu ohhtuceoughopghnohhtpghrvghplhihsehmnhhothdrnhgvtheqnecuggftrfgrthhtvg hrnhepkeefvdduteejvdefkeehieevuefgfefhteetveegffekffefteffvdelheduieet necuffhomhgrihhnpehgihhthhhusgdrtghomhenucfkphephedvrddukeekrdduiedurd dutdenucevlhhushhtvghrufhiiigvpedunecurfgrrhgrmhepmhgrihhlfhhrohhmpegu ohgpnhhothgprhgvphhlhiesmhhnohhtrdhnvght
X-ME-Proxy: <xmx:ptLlXl5L1USQ3sDDa4A4QbjPrm4Z3S2MFAT6cawA5tQhi9uPTDi3pQ> <xmx:ptLlXjfjj-YAGdKdTxB30jAgnW7NCGDMgjAaj0tLOvR4p4pWKidRtQ> <xmx:ptLlXpKU_t5LW6I7ea5CK1nXQK7ojUpJbXGqucxpZjeKfmAp70LJsA> <xmx:ptLlXiyQ6z6IxAuURE9k_qdNjZGKeEMECgMcbwJITYhpOpkNRrrnKw>
Received: from fv-az151.internal.cloudapp.net (unknown [52.188.161.10]) by mail.messagingengine.com (Postfix) with ESMTPA id 13CBB3060F09 for <quic@ietf.org>; Sun, 14 Jun 2020 03:32:54 -0400 (EDT)
Content-Type: multipart/alternative; boundary="===============2261701875098900325=="
MIME-Version: 1.0
From: Repository Activity Summary Bot <do_not_reply@mnot.net>
To: quic@ietf.org
Subject: Weekly github digest (QUIC Activity Summary)
Message-Id: <20200614073254.13CBB3060F09@mailuser.nyi.internal>
Date: Sun, 14 Jun 2020 03:32:54 -0400 (EDT)
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/UNm3P7FbsU5fk3SrvOnRGgXkq-o>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Sun, 14 Jun 2020 07:32:58 -0000

--===============2261701875098900325==
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain; charset="utf-8"; format="flowed"




Events without label "editorial"

Issues
------
* quicwg/base-drafts (+1/-9/=F0=9F=92=AC3)
  1 issues created:
  - List all transport parameters in "Values of Transport Parameters for 0-=
RTT" section (by nharper)
    https://github.com/quicwg/base-drafts/issues/3755=20

  3 issues received 3 new comments:
  - #3709 Make SERVER_BUSY more generic (1 by martinthomson)
    https://github.com/quicwg/base-drafts/issues/3709 [-transport] [design]=
 [has-consensus]=20
  - #3661 Include epoch in the AAD or the nonce? (1 by LPardue)
    https://github.com/quicwg/base-drafts/issues/3661 [-tls] [design] [prop=
osal-ready]=20
  - #3526 QUIC PTO is too conservative, causing a measurable regression in =
tail latency (1 by martinthomson)
    https://github.com/quicwg/base-drafts/issues/3526 [-recovery] [design] =
[future-version] [has-consensus]=20

  9 issues closed:
  - HTTP: provide guidance on PING usage https://github.com/quicwg/base-dra=
fts/issues/3723 [-http] [-transport]=20
  - QUIC PTO is too conservative, causing a measurable regression in tail l=
atency https://github.com/quicwg/base-drafts/issues/3526 [-recovery] [desig=
n] [future-version] [has-consensus]=20
  - Changing the Default QUIC ACK Policy https://github.com/quicwg/base-dra=
fts/issues/3529 [-recovery] [-transport] [design] [has-consensus]=20
  - Text on ECN probing https://github.com/quicwg/base-drafts/issues/3585 [=
-transport] [design] [has-consensus]=20
  - 5-tuple routing and SPA https://github.com/quicwg/base-drafts/issues/36=
08 [-transport] [design] [has-consensus]=20
  - Forgery limits on packet protection https://github.com/quicwg/base-draf=
ts/issues/3619 [-tls] [design] [has-consensus]=20
  - Integrate QUIC DPLPMTUD text https://github.com/quicwg/base-drafts/issu=
es/3695 [-transport] [design] [has-consensus]=20
  - use TRANSPORT_PARAMETER_ERROR when authenticating connection IDs https:=
//github.com/quicwg/base-drafts/issues/3703 [-transport] [design] [has-cons=
ensus]=20
  - Make SERVER_BUSY more generic https://github.com/quicwg/base-drafts/iss=
ues/3709 [-transport] [design] [has-consensus]=20

* quicwg/datagram (+0/-0/=F0=9F=92=AC5)
  1 issues received 5 new comments:
  - #3 Specify Max Payload Size instead of Max Frame Size (5 by DavidSchina=
zi, mikkelfj)
    https://github.com/quicwg/datagram/issues/3=20

* quicwg/load-balancers (+1/-0/=F0=9F=92=AC0)
  1 issues created:
  - Discuss uniqueness of config across load balancers (by martinduke)
    https://github.com/quicwg/load-balancers/issues/27=20



Pull requests
-------------
* quicwg/base-drafts (+9/-5/=F0=9F=92=AC7)
  9 pull requests submitted:
  - qpack: missing brackets around reference (by reschke)
    https://github.com/quicwg/base-drafts/pull/3761=20
  - Make usage of "out of order" and "out-of-order" consistent (by dtikhono=
v)
    https://github.com/quicwg/base-drafts/pull/3760=20
  - Use the same spelling of "trade-off" in two places (by dtikhonov)
    https://github.com/quicwg/base-drafts/pull/3759=20
  - http: expand "ALPN" on first use and ref RFC 7301 (by reschke)
    https://github.com/quicwg/base-drafts/pull/3758=20
  - http: no ABNF (by reschke)
    https://github.com/quicwg/base-drafts/pull/3757=20
  - Mention every Transport Parameter in section on values for 0-RTT (by nh=
arper)
    https://github.com/quicwg/base-drafts/pull/3756=20
  - transport: use "CONNECTION_REFUSED" consistently (by Lekensteyn)
    https://github.com/quicwg/base-drafts/pull/3754=20
  -  Update Initial salt, Retry keys, and samples (#3711) to changelog (by =
alagoutte)
    https://github.com/quicwg/base-drafts/pull/3752=20
  - Add issue template (by LPardue)
    https://github.com/quicwg/base-drafts/pull/3750=20

  7 pull requests received 7 new comments:
  - #3758 http: expand "ALPN" on first use and ref RFC 7301 (1 by reschke)
    https://github.com/quicwg/base-drafts/pull/3758=20
  - #3704 Revert "rename SERVER_BUSY to CONNECTION_REFUSED" (1 by martintho=
mson)
    https://github.com/quicwg/base-drafts/pull/3704=20
  - #3702 DPLPMTU merge tweaks (1 by martinthomson)
    https://github.com/quicwg/base-drafts/pull/3702 [-transport] [design]=20
  - #3623 Update ACK generation policy (1 by janaiyengar)
    https://github.com/quicwg/base-drafts/pull/3623 [-transport] [design]=20
  - #3501 Update ACK generation policy (1 by martinthomson)
    https://github.com/quicwg/base-drafts/pull/3501 [-transport] [design]=20
  - #3455 Add handling of all packets marked as CE (1 by martinthomson)
    https://github.com/quicwg/base-drafts/pull/3455 [-transport] [design]=20
  - #3454 In an unlikely event of emergency.... set ECN marks (1 by martint=
homson)
    https://github.com/quicwg/base-drafts/pull/3454 [-transport] [design]=20

  5 pull requests merged:
  - Add issue template
    https://github.com/quicwg/base-drafts/pull/3750=20
  - Scope of disable_active_migration
    https://github.com/quicwg/base-drafts/pull/3670 [-transport] [design]=20
  - Define an anti-forgery limit
    https://github.com/quicwg/base-drafts/pull/3620 [-tls] [design]=20
  - DPLPMTU merge tweaks
    https://github.com/quicwg/base-drafts/pull/3702 [-transport] [design]=20
  - use TRANSPORT_PARAMETER_ERROR when authenticating connection IDs
    https://github.com/quicwg/base-drafts/pull/3691 [-transport] [design]=20


Repositories tracked by this digest:
-----------------------------------
* https://github.com/quicwg/base-drafts
* https://github.com/quicwg/ops-drafts
* https://github.com/quicwg/datagram
* https://github.com/quicwg/load-balancers
* https://github.com/quicwg/version-negotiation

--===============2261701875098900325==
Content-Type: text/html; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable

<!doctype html>
<html lang=3D"en">
<head>
<meta charset=3D"utf-8">
<title>Weekly github digest (QUIC Activity Summary)</title>
<style>
body { font-family: Gotham, "Helvetica Neue", Helvetica, Arial, sans-serif;=
 font-size: 14px; }
h2 { margin-top: 3em; color: #A52A2A; font-style: italic; font-weight: norm=
al; }
h3 { margin-bottom:0; margin-top: 2em; font-size: 1.2em; }
h1+h2 { margin-top: 1em; }
a { color: #bb6219; text-decoration: none; }
li { margin-bottom: .35em; }
.repos { margin-bottom: 0; margin-top:0; line-height: 1.2; }
.new { color: red; }
.label { display: inline;
	padding: .2em .6em .3em;
	font-size: 75%;
	font-weight: 700;
	line-height: 1;
	color: #fff;
	text-align: center;
	white-space: nowrap;
	vertical-align: baseline;
	border-radius: .25em;
}
</style>
</head>

<body>
<h1>Sunday June 14, 2020</h1>

<p>Events without label "editorial"</p>

<h2>Issues</h2>

<h3>quicwg/base-drafts (+1/-9/=F0=9F=92=AC3)</h3>
  <p class=3D"new">1 issues created:</p>
  <ul>
  <li>#3755 <a href=3D"https://github.com/quicwg/base-drafts/issues/3755">L=
ist all transport parameters in &quot;Values of Transport Parameters for 0-=
RTT&quot; section</a> (by nharper) </li>
  </ul>

  <p>3 issues received 3 new comments:</p>
  <ul>
  <li>#3709 <a href=3D"https://github.com/quicwg/base-drafts/issues/3709">M=
ake SERVER_BUSY more generic</a> (1 by martinthomson) <span class=3D"label"=
 style=3D"background-color: #fbca04; color: #000000">-transport</span> <spa=
n class=3D"label" style=3D"background-color: #1d76db; color: #ffffff">desig=
n</span> <span class=3D"label" style=3D"background-color: #5319e7; color: #=
ffffff">has-consensus</span> </li>
 =20
  <li>#3661 <a href=3D"https://github.com/quicwg/base-drafts/issues/3661">I=
nclude epoch in the AAD or the nonce?</a> (1 by LPardue) <span class=3D"lab=
el" style=3D"background-color: #fbca04; color: #000000">-tls</span> <span c=
lass=3D"label" style=3D"background-color: #1d76db; color: #ffffff">design</=
span> <span class=3D"label" style=3D"background-color: #c5def5; color: #000=
000">proposal-ready</span> </li>
 =20
  <li>#3526 <a href=3D"https://github.com/quicwg/base-drafts/issues/3526">Q=
UIC PTO is too conservative, causing a measurable regression in tail latenc=
y</a> (1 by martinthomson) <span class=3D"label" style=3D"background-color:=
 #fbca04; color: #000000">-recovery</span> <span class=3D"label" style=3D"b=
ackground-color: #1d76db; color: #ffffff">design</span> <span class=3D"labe=
l" style=3D"background-color: #8e64e0; color: #ffffff">future-version</span=
> <span class=3D"label" style=3D"background-color: #5319e7; color: #ffffff"=
>has-consensus</span> </li>
  </ul>

  <p>9 issues closed:</p>
  <ul>
  <li>#3723 <a href=3D"https://github.com/quicwg/base-drafts/issues/3723">H=
TTP: provide guidance on PING usage</a> <span class=3D"label" style=3D"back=
ground-color: #fbca04; color: #000000">-http</span> <span class=3D"label" s=
tyle=3D"background-color: #fbca04; color: #000000">-transport</span> </li>
 =20
  <li>#3526 <a href=3D"https://github.com/quicwg/base-drafts/issues/3526">Q=
UIC PTO is too conservative, causing a measurable regression in tail latenc=
y</a> <span class=3D"label" style=3D"background-color: #fbca04; color: #000=
000">-recovery</span> <span class=3D"label" style=3D"background-color: #1d7=
6db; color: #ffffff">design</span> <span class=3D"label" style=3D"backgroun=
d-color: #8e64e0; color: #ffffff">future-version</span> <span class=3D"labe=
l" style=3D"background-color: #5319e7; color: #ffffff">has-consensus</span>=
 </li>
 =20
  <li>#3529 <a href=3D"https://github.com/quicwg/base-drafts/issues/3529">C=
hanging the Default QUIC ACK Policy</a> <span class=3D"label" style=3D"back=
ground-color: #fbca04; color: #000000">-recovery</span> <span class=3D"labe=
l" style=3D"background-color: #fbca04; color: #000000">-transport</span> <s=
pan class=3D"label" style=3D"background-color: #1d76db; color: #ffffff">des=
ign</span> <span class=3D"label" style=3D"background-color: #5319e7; color:=
 #ffffff">has-consensus</span> </li>
 =20
  <li>#3585 <a href=3D"https://github.com/quicwg/base-drafts/issues/3585">T=
ext on ECN probing</a> <span class=3D"label" style=3D"background-color: #fb=
ca04; color: #000000">-transport</span> <span class=3D"label" style=3D"back=
ground-color: #1d76db; color: #ffffff">design</span> <span class=3D"label" =
style=3D"background-color: #5319e7; color: #ffffff">has-consensus</span> </=
li>
 =20
  <li>#3608 <a href=3D"https://github.com/quicwg/base-drafts/issues/3608">5=
-tuple routing and SPA</a> <span class=3D"label" style=3D"background-color:=
 #fbca04; color: #000000">-transport</span> <span class=3D"label" style=3D"=
background-color: #1d76db; color: #ffffff">design</span> <span class=3D"lab=
el" style=3D"background-color: #5319e7; color: #ffffff">has-consensus</span=
> </li>
 =20
  <li>#3619 <a href=3D"https://github.com/quicwg/base-drafts/issues/3619">F=
orgery limits on packet protection</a> <span class=3D"label" style=3D"backg=
round-color: #fbca04; color: #000000">-tls</span> <span class=3D"label" sty=
le=3D"background-color: #1d76db; color: #ffffff">design</span> <span class=
=3D"label" style=3D"background-color: #5319e7; color: #ffffff">has-consensu=
s</span> </li>
 =20
  <li>#3695 <a href=3D"https://github.com/quicwg/base-drafts/issues/3695">I=
ntegrate QUIC DPLPMTUD text</a> <span class=3D"label" style=3D"background-c=
olor: #fbca04; color: #000000">-transport</span> <span class=3D"label" styl=
e=3D"background-color: #1d76db; color: #ffffff">design</span> <span class=
=3D"label" style=3D"background-color: #5319e7; color: #ffffff">has-consensu=
s</span> </li>
 =20
  <li>#3703 <a href=3D"https://github.com/quicwg/base-drafts/issues/3703">u=
se TRANSPORT_PARAMETER_ERROR when authenticating connection IDs</a> <span c=
lass=3D"label" style=3D"background-color: #fbca04; color: #000000">-transpo=
rt</span> <span class=3D"label" style=3D"background-color: #1d76db; color: =
#ffffff">design</span> <span class=3D"label" style=3D"background-color: #53=
19e7; color: #ffffff">has-consensus</span> </li>
 =20
  <li>#3709 <a href=3D"https://github.com/quicwg/base-drafts/issues/3709">M=
ake SERVER_BUSY more generic</a> <span class=3D"label" style=3D"background-=
color: #fbca04; color: #000000">-transport</span> <span class=3D"label" sty=
le=3D"background-color: #1d76db; color: #ffffff">design</span> <span class=
=3D"label" style=3D"background-color: #5319e7; color: #ffffff">has-consensu=
s</span> </li>
  </ul>

<h3>quicwg/datagram (+0/-0/=F0=9F=92=AC5)</h3>

  <p>1 issues received 5 new comments:</p>
  <ul>
  <li>#3 <a href=3D"https://github.com/quicwg/datagram/issues/3">Specify Ma=
x Payload Size instead of Max Frame Size</a> (5 by DavidSchinazi, mikkelfj)=
 </li>
  </ul>


<h3>quicwg/load-balancers (+1/-0/=F0=9F=92=AC0)</h3>
  <p class=3D"new">1 issues created:</p>
  <ul>
  <li>#27 <a href=3D"https://github.com/quicwg/load-balancers/issues/27">Di=
scuss uniqueness of config across load balancers</a> (by martinduke) </li>
  </ul>





<h2>Pull requests</h2>
<h3>quicwg/base-drafts (+9/-5/=F0=9F=92=AC7)</h3>
  <p class=3D"new">9 pull requests submitted:</p>
  <ul>
  <li>#3761 <a href=3D"https://github.com/quicwg/base-drafts/pull/3761">qpa=
ck: missing brackets around reference</a> (by reschke) </li>
 =20
  <li>#3760 <a href=3D"https://github.com/quicwg/base-drafts/pull/3760">Mak=
e usage of &quot;out of order&quot; and &quot;out-of-order&quot; consistent=
</a> (by dtikhonov) </li>
 =20
  <li>#3759 <a href=3D"https://github.com/quicwg/base-drafts/pull/3759">Use=
 the same spelling of &quot;trade-off&quot; in two places</a> (by dtikhonov=
) </li>
 =20
  <li>#3758 <a href=3D"https://github.com/quicwg/base-drafts/pull/3758">htt=
p: expand &quot;ALPN&quot; on first use and ref RFC 7301</a> (by reschke) <=
/li>
 =20
  <li>#3757 <a href=3D"https://github.com/quicwg/base-drafts/pull/3757">htt=
p: no ABNF</a> (by reschke) </li>
 =20
  <li>#3756 <a href=3D"https://github.com/quicwg/base-drafts/pull/3756">Men=
tion every Transport Parameter in section on values for 0-RTT</a> (by nharp=
er) </li>
 =20
  <li>#3754 <a href=3D"https://github.com/quicwg/base-drafts/pull/3754">tra=
nsport: use &quot;CONNECTION_REFUSED&quot; consistently</a> (by Lekensteyn)=
 </li>
 =20
  <li>#3752 <a href=3D"https://github.com/quicwg/base-drafts/pull/3752"> Up=
date Initial salt, Retry keys, and samples (#3711) to changelog</a> (by ala=
goutte) </li>
 =20
  <li>#3750 <a href=3D"https://github.com/quicwg/base-drafts/pull/3750">Add=
 issue template</a> (by LPardue) </li>
  </ul>

  <p>7 pull requests received 7 new comments:</p>
  <ul>
  <li>#3758 <a href=3D"https://github.com/quicwg/base-drafts/pull/3758">htt=
p: expand &quot;ALPN&quot; on first use and ref RFC 7301</a> (1 by reschke)=
 </li>
 =20
  <li>#3704 <a href=3D"https://github.com/quicwg/base-drafts/pull/3704">Rev=
ert &quot;rename SERVER_BUSY to CONNECTION_REFUSED&quot;</a> (1 by martinth=
omson) </li>
 =20
  <li>#3702 <a href=3D"https://github.com/quicwg/base-drafts/pull/3702">DPL=
PMTU merge tweaks</a> (1 by martinthomson) <span class=3D"label" style=3D"b=
ackground-color: #fbca04; color: #000000">-transport</span> <span class=3D"=
label" style=3D"background-color: #1d76db; color: #ffffff">design</span> </=
li>
 =20
  <li>#3623 <a href=3D"https://github.com/quicwg/base-drafts/pull/3623">Upd=
ate ACK generation policy</a> (1 by janaiyengar) <span class=3D"label" styl=
e=3D"background-color: #fbca04; color: #000000">-transport</span> <span cla=
ss=3D"label" style=3D"background-color: #1d76db; color: #ffffff">design</sp=
an> </li>
 =20
  <li>#3501 <a href=3D"https://github.com/quicwg/base-drafts/pull/3501">Upd=
ate ACK generation policy</a> (1 by martinthomson) <span class=3D"label" st=
yle=3D"background-color: #fbca04; color: #000000">-transport</span> <span c=
lass=3D"label" style=3D"background-color: #1d76db; color: #ffffff">design</=
span> </li>
 =20
  <li>#3455 <a href=3D"https://github.com/quicwg/base-drafts/pull/3455">Add=
 handling of all packets marked as CE</a> (1 by martinthomson) <span class=
=3D"label" style=3D"background-color: #fbca04; color: #000000">-transport</=
span> <span class=3D"label" style=3D"background-color: #1d76db; color: #fff=
fff">design</span> </li>
 =20
  <li>#3454 <a href=3D"https://github.com/quicwg/base-drafts/pull/3454">In =
an unlikely event of emergency.... set ECN marks</a> (1 by martinthomson) <=
span class=3D"label" style=3D"background-color: #fbca04; color: #000000">-t=
ransport</span> <span class=3D"label" style=3D"background-color: #1d76db; c=
olor: #ffffff">design</span> </li>
  </ul>

  <p>5 pull requests merged:</p>
  <ul>
  <li>#3750 <a href=3D"https://github.com/quicwg/base-drafts/pull/3750">Add=
 issue template</a> </li>
 =20
  <li>#3670 <a href=3D"https://github.com/quicwg/base-drafts/pull/3670">Sco=
pe of disable_active_migration</a> <span class=3D"label" style=3D"backgroun=
d-color: #fbca04; color: #">-transport</span> <span class=3D"label" style=
=3D"background-color: #1d76db; color: #">design</span> </li>
 =20
  <li>#3620 <a href=3D"https://github.com/quicwg/base-drafts/pull/3620">Def=
ine an anti-forgery limit</a> <span class=3D"label" style=3D"background-col=
or: #fbca04; color: #">-tls</span> <span class=3D"label" style=3D"backgroun=
d-color: #1d76db; color: #">design</span> </li>
 =20
  <li>#3702 <a href=3D"https://github.com/quicwg/base-drafts/pull/3702">DPL=
PMTU merge tweaks</a> <span class=3D"label" style=3D"background-color: #fbc=
a04; color: #">-transport</span> <span class=3D"label" style=3D"background-=
color: #1d76db; color: #">design</span> </li>
 =20
  <li>#3691 <a href=3D"https://github.com/quicwg/base-drafts/pull/3691">use=
 TRANSPORT_PARAMETER_ERROR when authenticating connection IDs</a> <span cla=
ss=3D"label" style=3D"background-color: #fbca04; color: #">-transport</span=
> <span class=3D"label" style=3D"background-color: #1d76db; color: #">desig=
n</span> </li>
  </ul>


<h2>Repositories tracked by this digest:</h2>
<ul class=3D"repos">
  <li><a href=3D"https://github.com/quicwg/base-drafts">https://github.com/=
quicwg/base-drafts</a></li>
  <li><a href=3D"https://github.com/quicwg/ops-drafts">https://github.com/q=
uicwg/ops-drafts</a></li>
  <li><a href=3D"https://github.com/quicwg/datagram">https://github.com/qui=
cwg/datagram</a></li>
  <li><a href=3D"https://github.com/quicwg/load-balancers">https://github.c=
om/quicwg/load-balancers</a></li>
  <li><a href=3D"https://github.com/quicwg/version-negotiation">https://git=
hub.com/quicwg/version-negotiation</a></li>
  </ul>
</body>
</html>

--===============2261701875098900325==--


From nobody Sun Jun 14 13:25:16 2020
Return-Path: <watsonbladd@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 96E103A1287 for <quic@ietfa.amsl.com>; Sun, 14 Jun 2020 13:25:14 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.099
X-Spam-Level: 
X-Spam-Status: No, score=-2.099 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Av9OG3aDPVPy for <quic@ietfa.amsl.com>; Sun, 14 Jun 2020 13:25:13 -0700 (PDT)
Received: from mail-lf1-x12a.google.com (mail-lf1-x12a.google.com [IPv6:2a00:1450:4864:20::12a]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 5FEC83A1285 for <quic@ietf.org>; Sun, 14 Jun 2020 13:25:13 -0700 (PDT)
Received: by mail-lf1-x12a.google.com with SMTP id u16so8303859lfl.8 for <quic@ietf.org>; Sun, 14 Jun 2020 13:25:13 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:from:date:message-id:subject:to; bh=kLjrGlPwepj9bcTAjbnujvRFJbV6XWnBZKkVhE/TbJI=; b=cvdGufi1k5eqrPjtovwDF2yk9kXgDRVEan4C1jBsVgfflAyFcw7pMbR54ouHF0btu1 reTcZ3o3Bx7RXs4mzKk8QAXm5KSFUVF1Vyk6S8zjp320FT46RtY3i3LSXCzBcKq+kRhh deml/Xkh1b2+s+Ub5j26Pr2ioHGbd3+X4Azt9XAkqV37N6laCQlSS+P3rBcKAPcO5n0h 3xZ455KZeK85rMkxbGNTNmwpRJ7oqiWm/blswnyRgrcgFrk0LCuoETLcYBFjWQcXVp7j WXIkulQXfpAQQ6dBkfTn26X9r7YrHF+HynXbIXmauOjRUKfTK7br5GUa6J8MszrEbVbg g0DA==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:from:date:message-id:subject:to; bh=kLjrGlPwepj9bcTAjbnujvRFJbV6XWnBZKkVhE/TbJI=; b=nQVcUnYXE95V/uVZ2IYc/c3muRA6Sal6ZEdMlIOUB6YPOdJrRUkaU9Va6kqPBx+0fO AnhTRnEKsDhmE/CVWWfvxWI/G3CpHKsiNvlV8L49X3xRviUH5EoB3KwdGOK1hJWmDlfF nj2y8oMT1lQGRgVfG9vJERoDm9PPzkgMUR6Jbbo4c1lPCb7zrG/qS4V+U6lKYS4893aP tp+miaWNFYDf+x6eMw+Txcomx5PzqMOIEiJMXkLNJhB6pYnwFr8MTgysHrBHwxzEJ9bJ ec/OXKtP6TOS9ZGLSF2Lmtfvsf9a7OGwI9N8FlKzyE75zrH/sFm+OrjO3gvOKy4zwwWf mV7w==
X-Gm-Message-State: AOAM532nn9VvVCm6EqM84htGjZ51QyZsDw1rwwkJ7eNWPVuLOYLT/2h9 YEzs85aVoFAPp8m+uHAW/P+7S4BE1fKVvEcxz19Ygv64
X-Google-Smtp-Source: ABdhPJwhHXg7MqT9REO60h+Ze9/gLzOTnGtbES6zHR4An2O6iGDSYXQJP9IbPLbEgKnBldpIDg92xN1DOqNx5IVOhmU=
X-Received: by 2002:ac2:5324:: with SMTP id f4mr12063943lfh.209.1592166311074;  Sun, 14 Jun 2020 13:25:11 -0700 (PDT)
MIME-Version: 1.0
From: Watson Ladd <watsonbladd@gmail.com>
Date: Sun, 14 Jun 2020 16:24:59 -0400
Message-ID: <CACsn0c=+P4vRpOHw-MD4NxU5nVPKORBHoNPYnxxtk=fk0OCGiA@mail.gmail.com>
Subject: Why not exporters?
To: IETF QUIC WG <quic@ietf.org>
Content-Type: text/plain; charset="UTF-8"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/R8Ch-i7KJ6TT5TzL5eSjsklebS0>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Sun, 14 Jun 2020 20:25:15 -0000

Dear QUIC wg,

I'm afraid I missed the discussion of using exporters in the TLS QUIC
draft. Exporters would assure the domain separation from TLS, and have
an IANA registry to maintain that separation across versions. There
doesn't seem to be a good reason not to do it.


Sincerely,
Watson Ladd


From nobody Sun Jun 14 16:37:48 2020
Return-Path: <mt@lowentropy.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 65C723A0544 for <quic@ietfa.amsl.com>; Sun, 14 Jun 2020 16:37:46 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.098
X-Spam-Level: 
X-Spam-Status: No, score=-2.098 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_MSPIKE_H3=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=lowentropy.net header.b=UZZpCz1m; dkim=pass (2048-bit key) header.d=messagingengine.com header.b=i7l8RbLp
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id tmpp9WDvtaOY for <quic@ietfa.amsl.com>; Sun, 14 Jun 2020 16:37:45 -0700 (PDT)
Received: from out1-smtp.messagingengine.com (out1-smtp.messagingengine.com [66.111.4.25]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 285263A0542 for <quic@ietf.org>; Sun, 14 Jun 2020 16:37:45 -0700 (PDT)
Received: from compute2.internal (compute2.nyi.internal [10.202.2.42]) by mailout.nyi.internal (Postfix) with ESMTP id 4E5085C00C7 for <quic@ietf.org>; Sun, 14 Jun 2020 19:37:44 -0400 (EDT)
Received: from imap2 ([10.202.2.52]) by compute2.internal (MEProxy); Sun, 14 Jun 2020 19:37:44 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=lowentropy.net; h=mime-version:message-id:in-reply-to:references:date:from:to :subject:content-type; s=fm2; bh=FgVqolhZof0czREqKcWLIoELXx+nz7+ lxB27Zj2CzbA=; b=UZZpCz1m8A/a2672WLVZJtXb3paAyKT8Nf6OLRNW06U5GbJ PWr3Q1y/JMH3uoZSlBSmaas1szEaULqBbTxCTWAoeZHP3247j6sLJPJjH/t0mxhJ TkWYwPxmD3EH9Y17S12uzpg3sZpYCnNlFyZdHxpU+JeSLN/2MQgeAkVPf36hAWtd NJTaAuohcWysw0x54LFC4ZFKemGsk5dTzHBhL2PE/a0Pbpw2SorbaYVGHs4rDLNU jJsR/KaKOQBqkUnmEYP6z5BBvLZQbLBAtZg0+woWZMCZn93HdtXhJ2+xG7AXbccw Ex/AxsCG/9lo+blnBn4xM4Az/RxUHK4hNeLVxVw==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d= messagingengine.com; h=content-type:date:from:in-reply-to :message-id:mime-version:references:subject:to:x-me-proxy :x-me-proxy:x-me-sender:x-me-sender:x-sasl-enc; s=fm3; bh=FgVqol hZof0czREqKcWLIoELXx+nz7+lxB27Zj2CzbA=; b=i7l8RbLpytQMstBoHyabHv QAEMngoWFpNDAIF1XFPIAxKghWqwmUC+VoDJRcGSS9bJQkDmFZTur1UO7wuG6Tji kWIneRF0GuGgb9LkYJ0sfi7r5TUAj0iR5JHili7KAgD8p/3Ow+JTtPMgRTmI+nJu 99EHZHgjuXCfkc2m9WzuICAGi1g6QdAYBgtnRe5iTw6qVcDxmShlDiCvzvQv3r+a 5dB/W1lgYFRN4reXQI5Wl8iiCqQEGMeu3A4fmI8g30qB1mja9pAjJebBa2k/wolU 6mOTbqd7JvN7027xCD6edKyuSNTCcXinlbEZk4J1xishy4tT/UoYkAiLi36xYrVA ==
X-ME-Sender: <xms:yLTmXs1p4EdaA-h5A25W1mvK546NqQ8PYQBIj0GiXRBZnnlSEgQRhw>
X-ME-Proxy-Cause: gggruggvucftvghtrhhoucdtuddrgeduhedrudeijedgvdeiucetufdoteggodetrfdotf fvucfrrhhofhhilhgvmecuhfgrshhtofgrihhlpdfqfgfvpdfurfetoffkrfgpnffqhgen uceurghilhhouhhtmecufedttdenucenucfjughrpefofgggkfgjfhffhffvufgtsehttd ertderredtnecuhfhrohhmpedfofgrrhhtihhnucfvhhhomhhsohhnfdcuoehmtheslhho figvnhhtrhhophihrdhnvghtqeenucggtffrrghtthgvrhhnpeekteeuieektdekleefke evhfekffevvdevgfekgfeluefgvdejjeegffeigedtjeenucevlhhushhtvghrufhiiigv pedtnecurfgrrhgrmhepmhgrihhlfhhrohhmpehmtheslhhofigvnhhtrhhophihrdhnvg ht
X-ME-Proxy: <xmx:yLTmXnFWjdCcQBOb1qJQyjq692FADzrQ-uySJprFZXJA3pqX1Lf24w> <xmx:yLTmXk6sY20fgiUMmW1iWn9xLHKvPKD9CODB5aBexrdC7Oy92ybc_A> <xmx:yLTmXl3dCeETcYAC21qGNN70Xl6FvhHpoPsowFcUbKKJRdbY_LkZRQ> <xmx:yLTmXvHg9ezTC83n8guRXLrgq_e_B2GX3BH9PEumlMSNBIFxIyflOw>
Received: by mailuser.nyi.internal (Postfix, from userid 501) id ED491E00A8; Sun, 14 Jun 2020 19:37:43 -0400 (EDT)
X-Mailer: MessagingEngine.com Webmail Interface
User-Agent: Cyrus-JMAP/3.3.0-dev0-525-ge8fa799-fm-20200609.001-ge8fa7990
Mime-Version: 1.0
Message-Id: <04a71698-968d-436b-9940-29afc9fea37d@www.fastmail.com>
In-Reply-To: <CACsn0c=+P4vRpOHw-MD4NxU5nVPKORBHoNPYnxxtk=fk0OCGiA@mail.gmail.com>
References: <CACsn0c=+P4vRpOHw-MD4NxU5nVPKORBHoNPYnxxtk=fk0OCGiA@mail.gmail.com>
Date: Mon, 15 Jun 2020 09:37:23 +1000
From: "Martin Thomson" <mt@lowentropy.net>
To: quic@ietf.org
Subject: Re: Why not exporters?
Content-Type: text/plain
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/ZVOfoKoPCurIxDZi0owOxOQa4Iw>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Sun, 14 Jun 2020 23:37:46 -0000

On Mon, Jun 15, 2020, at 06:24, Watson Ladd wrote:
> I'm afraid I missed the discussion of using exporters in the TLS QUIC
> draft. Exporters would assure the domain separation from TLS, and have
> an IANA registry to maintain that separation across versions. There
> doesn't seem to be a good reason not to do it.

Hi Watson,

We used exporters in an earlier version of the specification.  Indeed, the introduction of the early exporter in TLS 1.3 was - at least in part - in support of providing exporter keys that QUIC could use for 0-RTT.

However, when we moved to more closely following the TLS key schedule for packet protection, we decided to use TLS secrets rather than define a Handshake key exporter.  In the current design, those secrets are not used by TLS.

Domain separation is now achieved at the leaf. Until key updates occur, QUIC packet protection keys are derived from the same secrets that TLS uses for record protection, but the labels used for deriving the individual values (key, iv, hp) use a different label.  This ensures that keys cannot be synchronized.  That is in addition to the handshake being different due to containing a QUIC-specific extension and the guarantee provided by TLS that different connections have different keys. It's not as good as a clean separation might be, but there are several layers of defense.

Hope that helps,


From nobody Wed Jun 17 17:16:08 2020
Return-Path: <martin.h.duke@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 0E4CE3A07A6 for <quic@ietfa.amsl.com>; Wed, 17 Jun 2020 17:16:06 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=unavailable autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id HxCpDr4Of_gx for <quic@ietfa.amsl.com>; Wed, 17 Jun 2020 17:16:04 -0700 (PDT)
Received: from mail-io1-xd34.google.com (mail-io1-xd34.google.com [IPv6:2607:f8b0:4864:20::d34]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 7DF673A07A5 for <quic@ietf.org>; Wed, 17 Jun 2020 17:16:04 -0700 (PDT)
Received: by mail-io1-xd34.google.com with SMTP id u13so5043587iol.10 for <quic@ietf.org>; Wed, 17 Jun 2020 17:16:04 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=/ojIUdKSGVERBw0J90WiAOag/iigwn+Ud6qmNa/2E30=; b=pe+eSgdv/plu9JRdASzrF7bL8UX7JIQQZVoas9rEs8YOs8uBf8ClibUuDDT0SFf1kE 7fn7l6l5NW/cpmT0HAnx3nTlBU8avpA5vs6rOvD0pLFXsnkJwWBNpmD6hTPXw8AKpBGD CFIPJIberkRU6mu/dlpPwrp3mnjkPWkFIh2HzgEMhr0RNfHnIQ1gb7n87LtIJsoId2yO n3mlKKKy8IkRay2QfpwHC8Kr1FCIEHX/nsmXMoUxEgPbrqfYPVo1wN9p5PfeFRN+UHVc oc2PvdW4ofV+GloEa6jKpHBxpopa1AnJlXRUGAKdmKh9UWeN0kmjZCkzclZprVMYXoF2 kfEQ==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=/ojIUdKSGVERBw0J90WiAOag/iigwn+Ud6qmNa/2E30=; b=PgAxpUOGqjD1colUHfP0XUHwEHoQ3IRlgBabmBkVMjmTRWMeObGwJcVetW7mXLkdYT FmZtK5yaZwg60i2FOVJAN9rzDpp2Pqr+3nSF7H5idwFXbuq7Ja8YiR+99bWdU2XWTZ3O tF25rsUT5nXDtHzRSgC5k2PDk9/dR97lJoo7TcCDtS9NIq16GLhCltVxUa8c732uiMR3 /qi++Cg7TG1eNWyGPMNm76Y6EcK94i+3k3tr49W9Psn/Q4InWPqZl9IQmmRo/+u3R7uo cf/diyqfj8z5AIHq+epRnMu2CGnPVnuISKG5Fga6QdQBd+DLH9lybJzlHWPmXPMQaIo2 d4Vw==
X-Gm-Message-State: AOAM532ZOUUVxu4jRQk/Zx16Ur9tieJX4na1JWbJ8fazTEtM3tlecwBc wFeWVoFlrxKw5l+BLqhGIsYE5tPIiSYRD8/c7UeIbhqj
X-Google-Smtp-Source: ABdhPJwfGDYC2pZk61A2NlUeyhL6K0cf0CpcP3KuF19V7Aoo4x+Hj9VKlIIbtmgrjxv8xXTZRi/RXzEif7Larqriymo=
X-Received: by 2002:a02:3908:: with SMTP id l8mr1516012jaa.121.1592439363497;  Wed, 17 Jun 2020 17:16:03 -0700 (PDT)
MIME-Version: 1.0
References: <CAM4esxQBqfrz24riPQA_VGKcGp_TzW0pqb97KfFMtNdW9pUfDg@mail.gmail.com> <833A693C-62E6-4889-9954-FCE65A839A7C@eggert.org> <CAKcm_gPMO2DtqvKucqVw0zDjSniSOmFD4p1Tp4YLjr9WSWdEUw@mail.gmail.com> <CAJU8_nUN42gGmQof24XD9-EjXedyzcarDyRP8MGe1qW-BZ=+Aw@mail.gmail.com> <9cd91c24-c730-22a4-7aa0-baf61613b3ce@huitema.net> <f4922cdb59014202900de44cc5fea0ff@usma1ex-dag1mb5.msg.corp.akamai.com>
In-Reply-To: <f4922cdb59014202900de44cc5fea0ff@usma1ex-dag1mb5.msg.corp.akamai.com>
From: Martin Duke <martin.h.duke@gmail.com>
Date: Wed, 17 Jun 2020 17:15:52 -0700
Message-ID: <CAM4esxQvwkTvpUcu6-+W5zWo22m-R1jvN7DcCpXfuw8Hb55qsw@mail.gmail.com>
Subject: Re: Is the invariants draft really standards track?
To: "Lubashev, Igor" <ilubashe@akamai.com>
Cc: Christian Huitema <huitema@huitema.net>, Kyle Rose <krose@krose.org>,  Ian Swett <ianswett=40google.com@dmarc.ietf.org>, Lars Eggert <lars@eggert.org>, IETF QUIC WG <quic@ietf.org>, Jared Mauch <jared@puck.nether.net>
Content-Type: multipart/alternative; boundary="00000000000096764d05a850ae80"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/Tk1VCOboFJKmb1XjYoQIzPesnlw>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 18 Jun 2020 00:16:06 -0000

--00000000000096764d05a850ae80
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Hi Igor, you might want to check out the "Retry Services" bit of the
QUIC-LB draft. This has something to do with the DDoS use case you discuss.

On Wed, May 27, 2020 at 9:07 AM Lubashev, Igor <ilubashe@akamai.com> wrote:

> I=E2=80=99m working on a manageability draft PR for this (how to rate lim=
it UDP to
> reduce disruption to QUIC if you have to rate limit UDP).  ETA end of the
> week (if I do not get pulled into something again).
>
>
>
> The relevant observation is that DDoS with UDP that is indistinguishable
> from QUIC will happen.  UDP is already the most prevalent DDoS vector,
> since it is easy for a compromised non-admin app to send a flood of huge
> UDP packets (with TCP you get throttled by the congestion controller).  S=
o
> there WILL be DDoS protection devices out there to try to mitigate the
> problem, possibly by observing both directions of the flow and deciding
> whether a packet belongs to a valid flow or not.
>
>
>
> Since such middle boxes will be created, the more explicit and normative
> Invariants are about what one can expect, the less such middle boxes may
> decide for themselves.  For example (I did not think long about it), if
> some elements of path validation could land into Invariants (roughly, =E2=
=80=9Cno
> more than X packets/bytes can be sent on a new path w/o a return packet=
=E2=80=9D),
> a DDoS middle box may use this fact and active connection migration might
> still have a chance during an attack (NAT rebinding could be linked by DD=
oS
> boxes to an old connection via unchanged CID).
>
>
>
>    - Igor
>
>
>
>
>
> *From:* Christian Huitema <huitema@huitema.net>
> *Sent:* Wednesday, May 27, 2020 11:34 AM
>
> On 5/27/2020 8:28 AM, Kyle Rose wrote:
>
> On Wed, May 27, 2020 at 10:34 AM Ian Swett <ianswett=3D
> 40google.com@dmarc.ietf.org> wrote:
>
> I was agreeing with MT, but I'm happy to see some more MUSTs added if
> people feel that'd be helpful.
>
>
>
> Coincidentally, we were just talking about this internally at Akamai
> yesterday. IMO, an invariants document isn't really helpful if it isn't
> normative, and for it to be normative it (or a related practices doc for
> operators) really needs to spell out clear boundaries for operators with
> MUSTs..
>
>
>
> The example that came up yesterday was around operators filtering QUIC in
> the event of a DDoS: one recommendation based on some conversations going
> back at least to Prague 2019 was to hash packets on 4-tuple and filter
> those below a hash value chosen for a desired ingress limit instead of
> doing what most operators do with UDP today, which is to cap UDP throughp=
ut
> and just drop packets randomly or tail drop.
>
> Interesting. Did they consider using the CID, or a fraction of it? This
> looks entirely like the scenario for which we developed stateless retry.
>
>
>
> This recommendation certainly imposes some constraints on future protocol
> development that motivate new invariants: for instance, it would preclude
> sharding a connection across multiple source ports (not that there is
> necessarily a reason to do this; it's just an example). But more
> importantly, it goes beyond invariants: it's one among many practices
> compatible with the current set of invariants, some reasonable and some
> terrible.
>
> This would break the "preferred address" redirection. Preferred address
> migration may or may not be spelled out in the invariants.
>
>
>
> Operators are going to do things to QUIC traffic, so it would be good to
> offer them recommendations that are compatible with broad deployability.
>
>
>
> Yes, we do need the invariants for that.
>
> -- Christian Huitema
>

--00000000000096764d05a850ae80
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Hi Igor, you might want to check out the &quot;Retry Servi=
ces&quot; bit of the QUIC-LB draft. This has something to do with the DDoS =
use case you discuss.</div><br><div class=3D"gmail_quote"><div dir=3D"ltr" =
class=3D"gmail_attr">On Wed, May 27, 2020 at 9:07 AM Lubashev, Igor &lt;<a =
href=3D"mailto:ilubashe@akamai.com">ilubashe@akamai.com</a>&gt; wrote:<br><=
/div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;bo=
rder-left:1px solid rgb(204,204,204);padding-left:1ex">





<div lang=3D"EN-US">
<div class=3D"gmail-m_-3800542681916443443WordSection1">
<p class=3D"MsoNormal">I=E2=80=99m working on a manageability draft PR for =
this (how to rate limit UDP to reduce disruption to QUIC if you have to rat=
e limit UDP).=C2=A0 ETA end of the week (if I do not get pulled into someth=
ing again).<u></u><u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal">The relevant observation is that DDoS with UDP that =
is indistinguishable from QUIC will happen.=C2=A0 UDP is already the most p=
revalent DDoS vector, since it is easy for a compromised non-admin app to s=
end a flood of huge UDP packets (with TCP
 you get throttled by the congestion controller).=C2=A0 So there WILL be DD=
oS protection devices out there to try to mitigate the problem, possibly by=
 observing both directions of the flow and deciding whether a packet belong=
s to a valid flow or not.<u></u><u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal">Since such middle boxes will be created, the more ex=
plicit and normative Invariants are about what one can expect, the less suc=
h middle boxes may decide for themselves.=C2=A0 For example (I did not thin=
k long about it), if some elements of path
 validation could land into Invariants (roughly, =E2=80=9Cno more than X pa=
ckets/bytes can be sent on a new path w/o a return packet=E2=80=9D), a DDoS=
 middle box may use this fact and active connection migration might still h=
ave a chance during an attack (NAT rebinding could
 be linked by DDoS boxes to an old connection via unchanged CID).<u></u><u>=
</u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<ul style=3D"margin-top:0in" type=3D"disc">
<li class=3D"MsoNormal">Igor<u></u><u></u></li></ul>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<div style=3D"border-top:none;border-right:none;border-bottom:none;border-l=
eft:1.5pt solid blue;padding:0in 0in 0in 4pt">
<p class=3D"MsoNormal"><b>From:</b> Christian Huitema &lt;<a href=3D"mailto=
:huitema@huitema.net" target=3D"_blank">huitema@huitema.net</a>&gt; <br>
<b>Sent:</b> Wednesday, May 27, 2020 11:34 AM<br>
<br>
</p>
<div>
<p class=3D"MsoNormal">On 5/27/2020 8:28 AM, Kyle Rose wrote:<u></u><u></u>=
</p>
</div>
<blockquote style=3D"margin-top:5pt;margin-bottom:5pt">
<p class=3D"MsoNormal">On Wed, May 27, 2020 at 10:34 AM Ian Swett &lt;iansw=
ett=3D<a href=3D"mailto:40google.com@dmarc.ietf.org" target=3D"_blank">40go=
ogle.com@dmarc.ietf.org</a>&gt; wrote:<u></u><u></u></p>
<div>
<blockquote style=3D"border-top:none;border-right:none;border-bottom:none;b=
order-left:1pt solid rgb(204,204,204);padding:0in 0in 0in 6pt;margin-left:4=
.8pt;margin-right:0in">
<div>
<p class=3D"MsoNormal">I was agreeing with MT, but I&#39;m happy to=C2=A0se=
e some more MUSTs added if people feel that&#39;d be helpful.<u></u><u></u>=
</p>
</div>
</blockquote>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt"><u></u>=C2=A0<u></u><=
/span></p>
<div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt">Coincidentally, we we=
re just talking about this internally at Akamai yesterday. IMO, an invarian=
ts document isn&#39;t really helpful if it isn&#39;t normative, and for it =
to be normative it (or a related practices
 doc for operators) really needs to spell out clear boundaries for operator=
s with MUSTs..<u></u><u></u></span></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt"><u></u>=C2=A0<u></u><=
/span></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt">The example that came=
 up yesterday was around operators filtering QUIC in the event of a DDoS: o=
ne recommendation based on some conversations going back at least to Prague=
 2019 was to hash packets on 4-tuple
 and filter those below a hash value chosen for a desired ingress limit ins=
tead of doing what most operators do with UDP today, which is to cap UDP th=
roughput and just drop packets randomly or tail drop.<u></u><u></u></span><=
/p>
</div>
</div>
</div>
</div>
</blockquote>
<p>Interesting. Did they consider using the CID, or a fraction of it? This =
looks entirely like the scenario for which we developed stateless retry.<u>=
</u><u></u></p>
<blockquote style=3D"margin-top:5pt;margin-bottom:5pt">
<div>
<div>
<div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt"><u></u>=C2=A0<u></u><=
/span></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt">This recommendation c=
ertainly imposes some constraints on future protocol development that motiv=
ate new invariants: for instance, it would preclude sharding a connection a=
cross multiple source ports (not that
 there is necessarily a reason to do this; it&#39;s just an example). But m=
ore importantly, it goes beyond invariants: it&#39;s one among many practic=
es compatible with the current set of invariants, some reasonable and some =
terrible.<u></u><u></u></span></p>
</div>
</div>
</div>
</div>
</blockquote>
<p>This would break the &quot;preferred address&quot; redirection. Preferre=
d address migration may or may not be spelled out in the invariants.<u></u>=
<u></u></p>
<blockquote style=3D"margin-top:5pt;margin-bottom:5pt">
<div>
<div>
<div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt"><u></u>=C2=A0<u></u><=
/span></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt">Operators are going t=
o do things to QUIC traffic, so it would be good to offer them recommendati=
ons that are compatible with broad deployability.<u></u><u></u></span></p>
</div>
</div>
</div>
</div>
</blockquote>
<p><u></u>=C2=A0<u></u></p>
<p>Yes, we do need the invariants for that.<u></u><u></u></p>
<p>-- Christian Huitema<u></u><u></u></p>
</div>
</div>
</div>

</blockquote></div>

--00000000000096764d05a850ae80--


From nobody Wed Jun 17 17:17:08 2020
Return-Path: <martin.h.duke@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 9DB983A07AF for <quic@ietfa.amsl.com>; Wed, 17 Jun 2020 17:17:07 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id FXg9SZxUdST3 for <quic@ietfa.amsl.com>; Wed, 17 Jun 2020 17:17:05 -0700 (PDT)
Received: from mail-il1-x12a.google.com (mail-il1-x12a.google.com [IPv6:2607:f8b0:4864:20::12a]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id C85773A07A5 for <quic@ietf.org>; Wed, 17 Jun 2020 17:17:05 -0700 (PDT)
Received: by mail-il1-x12a.google.com with SMTP id b5so4142184iln.5 for <quic@ietf.org>; Wed, 17 Jun 2020 17:17:05 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=SJJvauAYKKKDk/0UjyDGqOOvhkWIGSs6w+RuK6gGFv4=; b=u96JxUmdM2p+2VmcU0Lk0HdccRu4VvnklFJ6gwUhZ5bvBjV5CuG5d6frpqCBu2isU4 G/Kk3LLc2sMK8kVkalWC9eEIvX+cK2HaW094O3WcxoZ+RKdYsc+tBHlYPzYeO+98wz4o /V6iLJpd9/a1di1fx7IJGlKfwV7BXF0rj00vxEjOsRoOnT7ue1VpV0k8x5JBY0TQtm1j USjLyHkt+LZnj+a1euqfU8k1qCFZqI4ohsY5IRaWGAqmvEoMFY3H36c/U7zM19UM9iSp L+nhkTaRMojdfgvnlxGUDdrqtdd911GSOnE3zWafjW9RKJHKOYTkCd0FFb6cbEaIhcOs WlRQ==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=SJJvauAYKKKDk/0UjyDGqOOvhkWIGSs6w+RuK6gGFv4=; b=kzj1IZFNL+6OF3P/6JCYTiS4mHgZ+HJOnkHq2kkivxnYNYY/9XbCQ2Vf8Wit+WyZFF fw5j0XQEkKWn0Hm/CiQAoc4A1xh9DE2DlZS0lXORQjiz4tW7koB1J3ybW21A0uwG4T/L FnhY25fT/y7DILfKzlxBMKwKAT+F38xc/xVZBYedpwflMxuiuWhQUlCQLfYASR96bsGl /R4x6CAfnJR3Im1L5mffJcXeY+WOcBiUDjN+uUoRutcxcTX+kELkERzRFVLpjLQfVPcH DhJOsRB0MiO2zW4LSiKVGmG25gXE/2qQkdpmyzM7915nZhi3V+Sk3FEvw9PSEagNi5SC PBmA==
X-Gm-Message-State: AOAM532ekF6J3qzy9On2P5gsxXiTAH2L/5R+WYm580JrpQMobfjJbbr+ T1U1cpr1WFMlu4MEBTVcBhngP0hEA+e6a/2SSUE3LWTX
X-Google-Smtp-Source: ABdhPJzj8EERVVHMclBXNLqPf81pX1db/FoPGgDslFtMFhxBaExwYfHbI8EtWgyRidH/KyYV935JIAbr/mmSai+qHkU=
X-Received: by 2002:a05:6e02:f44:: with SMTP id y4mr1591602ilj.237.1592439425199;  Wed, 17 Jun 2020 17:17:05 -0700 (PDT)
MIME-Version: 1.0
References: <CAM4esxQBqfrz24riPQA_VGKcGp_TzW0pqb97KfFMtNdW9pUfDg@mail.gmail.com> <833A693C-62E6-4889-9954-FCE65A839A7C@eggert.org> <CAKcm_gPMO2DtqvKucqVw0zDjSniSOmFD4p1Tp4YLjr9WSWdEUw@mail.gmail.com> <CAJU8_nUN42gGmQof24XD9-EjXedyzcarDyRP8MGe1qW-BZ=+Aw@mail.gmail.com> <9cd91c24-c730-22a4-7aa0-baf61613b3ce@huitema.net> <f4922cdb59014202900de44cc5fea0ff@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQvwkTvpUcu6-+W5zWo22m-R1jvN7DcCpXfuw8Hb55qsw@mail.gmail.com>
In-Reply-To: <CAM4esxQvwkTvpUcu6-+W5zWo22m-R1jvN7DcCpXfuw8Hb55qsw@mail.gmail.com>
From: Martin Duke <martin.h.duke@gmail.com>
Date: Wed, 17 Jun 2020 17:16:54 -0700
Message-ID: <CAM4esxTB9Pt4345-jUozUUE0i1Z-ccO2cwAaVEJGZAyvpsC4sg@mail.gmail.com>
Subject: Re: Is the invariants draft really standards track?
To: "Lubashev, Igor" <ilubashe@akamai.com>
Cc: Christian Huitema <huitema@huitema.net>, Kyle Rose <krose@krose.org>,  Ian Swett <ianswett=40google.com@dmarc.ietf.org>, Lars Eggert <lars@eggert.org>, IETF QUIC WG <quic@ietf.org>, Jared Mauch <jared@puck.nether.net>
Content-Type: multipart/alternative; boundary="00000000000043f30105a850b2e3"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/ClOoCD7ir-c1jKbMmTI_gX_9FSw>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 18 Jun 2020 00:17:08 -0000

--00000000000043f30105a850b2e3
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

filed issue #3773 <https://github.com/quicwg/base-drafts/issues/3773> and
PR #3774 <https://github.com/quicwg/base-drafts/pull/3774>.

On Wed, Jun 17, 2020 at 5:15 PM Martin Duke <martin.h.duke@gmail.com> wrote=
:

> Hi Igor, you might want to check out the "Retry Services" bit of the
> QUIC-LB draft. This has something to do with the DDoS use case you discus=
s.
>
> On Wed, May 27, 2020 at 9:07 AM Lubashev, Igor <ilubashe@akamai.com>
> wrote:
>
>> I=E2=80=99m working on a manageability draft PR for this (how to rate li=
mit UDP
>> to reduce disruption to QUIC if you have to rate limit UDP).  ETA end of
>> the week (if I do not get pulled into something again).
>>
>>
>>
>> The relevant observation is that DDoS with UDP that is indistinguishable
>> from QUIC will happen.  UDP is already the most prevalent DDoS vector,
>> since it is easy for a compromised non-admin app to send a flood of huge
>> UDP packets (with TCP you get throttled by the congestion controller).  =
So
>> there WILL be DDoS protection devices out there to try to mitigate the
>> problem, possibly by observing both directions of the flow and deciding
>> whether a packet belongs to a valid flow or not.
>>
>>
>>
>> Since such middle boxes will be created, the more explicit and normative
>> Invariants are about what one can expect, the less such middle boxes may
>> decide for themselves.  For example (I did not think long about it), if
>> some elements of path validation could land into Invariants (roughly, =
=E2=80=9Cno
>> more than X packets/bytes can be sent on a new path w/o a return packet=
=E2=80=9D),
>> a DDoS middle box may use this fact and active connection migration migh=
t
>> still have a chance during an attack (NAT rebinding could be linked by D=
DoS
>> boxes to an old connection via unchanged CID).
>>
>>
>>
>>    - Igor
>>
>>
>>
>>
>>
>> *From:* Christian Huitema <huitema@huitema.net>
>> *Sent:* Wednesday, May 27, 2020 11:34 AM
>>
>> On 5/27/2020 8:28 AM, Kyle Rose wrote:
>>
>> On Wed, May 27, 2020 at 10:34 AM Ian Swett <ianswett=3D
>> 40google.com@dmarc.ietf.org> wrote:
>>
>> I was agreeing with MT, but I'm happy to see some more MUSTs added if
>> people feel that'd be helpful.
>>
>>
>>
>> Coincidentally, we were just talking about this internally at Akamai
>> yesterday. IMO, an invariants document isn't really helpful if it isn't
>> normative, and for it to be normative it (or a related practices doc for
>> operators) really needs to spell out clear boundaries for operators with
>> MUSTs..
>>
>>
>>
>> The example that came up yesterday was around operators filtering QUIC i=
n
>> the event of a DDoS: one recommendation based on some conversations goin=
g
>> back at least to Prague 2019 was to hash packets on 4-tuple and filter
>> those below a hash value chosen for a desired ingress limit instead of
>> doing what most operators do with UDP today, which is to cap UDP through=
put
>> and just drop packets randomly or tail drop.
>>
>> Interesting. Did they consider using the CID, or a fraction of it? This
>> looks entirely like the scenario for which we developed stateless retry.
>>
>>
>>
>> This recommendation certainly imposes some constraints on future protoco=
l
>> development that motivate new invariants: for instance, it would preclud=
e
>> sharding a connection across multiple source ports (not that there is
>> necessarily a reason to do this; it's just an example). But more
>> importantly, it goes beyond invariants: it's one among many practices
>> compatible with the current set of invariants, some reasonable and some
>> terrible.
>>
>> This would break the "preferred address" redirection. Preferred address
>> migration may or may not be spelled out in the invariants.
>>
>>
>>
>> Operators are going to do things to QUIC traffic, so it would be good to
>> offer them recommendations that are compatible with broad deployability.
>>
>>
>>
>> Yes, we do need the invariants for that.
>>
>> -- Christian Huitema
>>
>

--00000000000043f30105a850b2e3
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">filed issue <a href=3D"https://github.com/quicwg/base-draf=
ts/issues/3773">#3773</a> and PR <a href=3D"https://github.com/quicwg/base-=
drafts/pull/3774">#3774</a>.</div><br><div class=3D"gmail_quote"><div dir=
=3D"ltr" class=3D"gmail_attr">On Wed, Jun 17, 2020 at 5:15 PM Martin Duke &=
lt;<a href=3D"mailto:martin.h.duke@gmail.com">martin.h.duke@gmail.com</a>&g=
t; wrote:<br></div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0p=
x 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div d=
ir=3D"ltr">Hi Igor, you might want to check out the &quot;Retry Services&qu=
ot; bit of the QUIC-LB draft. This has something to do with the DDoS use ca=
se you discuss.</div><br><div class=3D"gmail_quote"><div dir=3D"ltr" class=
=3D"gmail_attr">On Wed, May 27, 2020 at 9:07 AM Lubashev, Igor &lt;<a href=
=3D"mailto:ilubashe@akamai.com" target=3D"_blank">ilubashe@akamai.com</a>&g=
t; wrote:<br></div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0p=
x 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">





<div lang=3D"EN-US">
<div>
<p class=3D"MsoNormal">I=E2=80=99m working on a manageability draft PR for =
this (how to rate limit UDP to reduce disruption to QUIC if you have to rat=
e limit UDP).=C2=A0 ETA end of the week (if I do not get pulled into someth=
ing again).<u></u><u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal">The relevant observation is that DDoS with UDP that =
is indistinguishable from QUIC will happen.=C2=A0 UDP is already the most p=
revalent DDoS vector, since it is easy for a compromised non-admin app to s=
end a flood of huge UDP packets (with TCP
 you get throttled by the congestion controller).=C2=A0 So there WILL be DD=
oS protection devices out there to try to mitigate the problem, possibly by=
 observing both directions of the flow and deciding whether a packet belong=
s to a valid flow or not.<u></u><u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal">Since such middle boxes will be created, the more ex=
plicit and normative Invariants are about what one can expect, the less suc=
h middle boxes may decide for themselves.=C2=A0 For example (I did not thin=
k long about it), if some elements of path
 validation could land into Invariants (roughly, =E2=80=9Cno more than X pa=
ckets/bytes can be sent on a new path w/o a return packet=E2=80=9D), a DDoS=
 middle box may use this fact and active connection migration might still h=
ave a chance during an attack (NAT rebinding could
 be linked by DDoS boxes to an old connection via unchanged CID).<u></u><u>=
</u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<ul style=3D"margin-top:0in" type=3D"disc">
<li class=3D"MsoNormal">Igor<u></u><u></u></li></ul>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<div style=3D"border-top:none;border-right:none;border-bottom:none;border-l=
eft:1.5pt solid blue;padding:0in 0in 0in 4pt">
<p class=3D"MsoNormal"><b>From:</b> Christian Huitema &lt;<a href=3D"mailto=
:huitema@huitema.net" target=3D"_blank">huitema@huitema.net</a>&gt; <br>
<b>Sent:</b> Wednesday, May 27, 2020 11:34 AM<br>
<br>
</p>
<div>
<p class=3D"MsoNormal">On 5/27/2020 8:28 AM, Kyle Rose wrote:<u></u><u></u>=
</p>
</div>
<blockquote style=3D"margin-top:5pt;margin-bottom:5pt">
<p class=3D"MsoNormal">On Wed, May 27, 2020 at 10:34 AM Ian Swett &lt;iansw=
ett=3D<a href=3D"mailto:40google.com@dmarc.ietf.org" target=3D"_blank">40go=
ogle.com@dmarc.ietf.org</a>&gt; wrote:<u></u><u></u></p>
<div>
<blockquote style=3D"border-top:none;border-right:none;border-bottom:none;b=
order-left:1pt solid rgb(204,204,204);padding:0in 0in 0in 6pt;margin-left:4=
.8pt;margin-right:0in">
<div>
<p class=3D"MsoNormal">I was agreeing with MT, but I&#39;m happy to=C2=A0se=
e some more MUSTs added if people feel that&#39;d be helpful.<u></u><u></u>=
</p>
</div>
</blockquote>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt"><u></u>=C2=A0<u></u><=
/span></p>
<div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt">Coincidentally, we we=
re just talking about this internally at Akamai yesterday. IMO, an invarian=
ts document isn&#39;t really helpful if it isn&#39;t normative, and for it =
to be normative it (or a related practices
 doc for operators) really needs to spell out clear boundaries for operator=
s with MUSTs..<u></u><u></u></span></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt"><u></u>=C2=A0<u></u><=
/span></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt">The example that came=
 up yesterday was around operators filtering QUIC in the event of a DDoS: o=
ne recommendation based on some conversations going back at least to Prague=
 2019 was to hash packets on 4-tuple
 and filter those below a hash value chosen for a desired ingress limit ins=
tead of doing what most operators do with UDP today, which is to cap UDP th=
roughput and just drop packets randomly or tail drop.<u></u><u></u></span><=
/p>
</div>
</div>
</div>
</div>
</blockquote>
<p>Interesting. Did they consider using the CID, or a fraction of it? This =
looks entirely like the scenario for which we developed stateless retry.<u>=
</u><u></u></p>
<blockquote style=3D"margin-top:5pt;margin-bottom:5pt">
<div>
<div>
<div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt"><u></u>=C2=A0<u></u><=
/span></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt">This recommendation c=
ertainly imposes some constraints on future protocol development that motiv=
ate new invariants: for instance, it would preclude sharding a connection a=
cross multiple source ports (not that
 there is necessarily a reason to do this; it&#39;s just an example). But m=
ore importantly, it goes beyond invariants: it&#39;s one among many practic=
es compatible with the current set of invariants, some reasonable and some =
terrible.<u></u><u></u></span></p>
</div>
</div>
</div>
</div>
</blockquote>
<p>This would break the &quot;preferred address&quot; redirection. Preferre=
d address migration may or may not be spelled out in the invariants.<u></u>=
<u></u></p>
<blockquote style=3D"margin-top:5pt;margin-bottom:5pt">
<div>
<div>
<div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt"><u></u>=C2=A0<u></u><=
/span></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt">Operators are going t=
o do things to QUIC traffic, so it would be good to offer them recommendati=
ons that are compatible with broad deployability.<u></u><u></u></span></p>
</div>
</div>
</div>
</div>
</blockquote>
<p><u></u>=C2=A0<u></u></p>
<p>Yes, we do need the invariants for that.<u></u><u></u></p>
<p>-- Christian Huitema<u></u><u></u></p>
</div>
</div>
</div>

</blockquote></div>
</blockquote></div>

--00000000000043f30105a850b2e3--


From nobody Thu Jun 18 01:24:00 2020
Return-Path: <kazuhooku@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C504E3A0F93 for <quic@ietfa.amsl.com>; Thu, 18 Jun 2020 01:23:58 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id gdeYzgJm8g9Z for <quic@ietfa.amsl.com>; Thu, 18 Jun 2020 01:23:57 -0700 (PDT)
Received: from mail-ed1-x52e.google.com (mail-ed1-x52e.google.com [IPv6:2a00:1450:4864:20::52e]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id D76473A0F91 for <quic@ietf.org>; Thu, 18 Jun 2020 01:23:56 -0700 (PDT)
Received: by mail-ed1-x52e.google.com with SMTP id k8so4190311edq.4 for <quic@ietf.org>; Thu, 18 Jun 2020 01:23:56 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:from:date:message-id:subject:to; bh=RB2LyGmEEbeJX2lFmECaRShcWyDjiH/Ic/oflYy7rpM=; b=THodY3Hua875Lt0cioeZw7/1c3hGi5g1afotkJx+1VAJ5ZR2TnmshEJmZQgPpgD76H GcE5gjoSzYm6C21/lV2/YfeSvVIkt3R4oowbLcB6SjaVFAlWHaHJ1xVJDboMrsfeqcxg hMwYhTh8diNQ6wjz78I9AxjJQaomCapU9aOiIpfQ6QlO5ZDtB+mJ/7WAJrcDSa7MZVNG 6WGP5HABmg40jqVgE+bTGYvSQzzxCILywJZ857uCf/3HKyreps4RP++OKSjJFBvtDbxD tVVFilyjCy4RBERe9tCs4Rjstfe1P57tfsg8zfBJYEONm36SPAfNhS8K+dkJJEfvyCty quEA==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:from:date:message-id:subject:to; bh=RB2LyGmEEbeJX2lFmECaRShcWyDjiH/Ic/oflYy7rpM=; b=LPWA5tsKCX1c3y2C9/6X/hbh5BVX9PNG9Y5KE6VbmNVCBA91uwXYgf9kDJeiHMoloU 98D1qen9VGd1+v9DM2EDoViWpnnzoRVZmTgIbBXuGydKVghbkYLiJj7hXROg58k61aQ6 EbV9vpefDogI41LxYZAvQEs8w1ZNoU+CvF46jgIvq8vhCRZebrcTYL0UUbJGHDUe+/JD yg+UnqZuoWh+ycEui7GLNPYoRC22m0Gfq+X2jEt5D2sX//azTbrp8wgUxL617XNq0aiP ecN3UW7K8aZk34tDhc7OFQzqQWt8nsHEAO05B7Cd7W+e0KDCiw77BCw5fZB/+Mi7UO1O d1yQ==
X-Gm-Message-State: AOAM530zPugqLNyU1NFQPKpeX0pv/H16PRzhU6dH+9+RZNxdr3D4of3m uANmLkz4X9syC6x7tZsH4AAVJt03n5wVqYiwKchC/snn
X-Google-Smtp-Source: ABdhPJwT8voZ0iunGWWzvxmsfqNL+NyrFPEZamp2yxSvqZAO7UgInXpC7uNMlLmzV9NsnNYu8mrBCC9gHjRXXcJw+Ls=
X-Received: by 2002:a05:6402:362:: with SMTP id s2mr2968770edw.337.1592468635103;  Thu, 18 Jun 2020 01:23:55 -0700 (PDT)
MIME-Version: 1.0
From: Kazuho Oku <kazuhooku@gmail.com>
Date: Thu, 18 Jun 2020 17:23:44 +0900
Message-ID: <CANatvzz8F1H=DXMkBEhmKHnYM-HVG48TS9KwY=OP881Txkcodw@mail.gmail.com>
Subject: AEAD and header encryption overhead in QUIC
To: IETF QUIC WG <quic@ietf.org>
Content-Type: multipart/alternative; boundary="0000000000004fb7db05a8577f73"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/OnBhC4gCosmlU3VKoFk_jJnpeXU>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 18 Jun 2020 08:23:59 -0000

--0000000000004fb7db05a8577f73
Content-Type: text/plain; charset="UTF-8"

Hello folks,

TL;DR: When existing crypto libraries are used, the AEAD cost of QUIC can
become somewhat greater than that of TLS over TCP, due to those crypto
libraries not optimized for handling short AEAD blocks. Crypto libraries
optimized for short AEAD blocks can tighten the gap. Also, crypto libraries
optimized for QUIC can combine AEAD operation and header protection vector
generation, effectively eliminating the cost of header protection.

A bit longer story:

Recently, when testing quicly [1] using a NIC with *hardware support* for
UDP GSO offload, we noticed that it uses +30% CPU cycles compared to TLS
over TCP. This was due to our underlying crypto not being optimized for
short AEAD blocks, and due to the call overhead of doing header protection
one packet at a time. [2]

We went on to implement our own AES-GCM library optimized for short AEAD
blocks (i.e. packet-sized), that also generates the header protection
vector while doing AEAD computation. By switching to the new crypto
library, the performance drawback of the crypto itself compared to large
AEAD blocks when down from 30% to 10% [3].

The overhead of header protection (relative to all the crypto cost) on the
send side is now between 0.3% to 1% when using packets sized around 1400
bytes to 1500 bytes. The ratio against the total transport cost is going to
be less than that, and in practice, I think that they are negligible. The
story might be a bit different on the receive-side due to the dependency
between header protection and AEAD, though I am sure there would also be
ways to optimize.

In terms of CPU cycles spent by the QUIC stack, the overhead compared to
TLS-over-TCP came down to 5%. quicly can now sustain the send rate of 5Gbps
while spending about 30% of CPU cycles of one modern Intel CPU core [2].

These numbers are observations from one TLS / QUIC stack doing
micro-benchmarks, though I tend to think that sharing performance numbers
lead to people having better expectations regarding the CPU cost of QUIC.
Some of us were also concerned about the cost of header protection (a.k.a.
packet number encryption) [4]. Hence posting here.

[1] https://github.com/h2o/quicly
[2] https://github.com/h2o/quicly/pull/359
[3] https://github.com/h2o/picotls/pull/310
[4] https://mailarchive.ietf.org/arch/msg/quic/Wy8JXwKyeYgWgRWizWDbyOLg0xc/

-- 
Kazuho Oku

--0000000000004fb7db05a8577f73
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div dir=3D"ltr"><div dir=3D"ltr"><div dir=3D"ltr"><div di=
r=3D"ltr">Hello folks,<div><br></div><div>TL;DR: When existing crypto libra=
ries are used, the AEAD cost of QUIC can become somewhat greater than that =
of TLS over TCP, due to those crypto libraries not optimized for handling s=
hort AEAD blocks. Crypto libraries optimized for=C2=A0short AEAD blocks can=
 tighten the gap. Also, crypto libraries optimized for QUIC can combine AEA=
D operation and header protection vector generation, effectively eliminatin=
g the cost of header protection.</div><div><br></div><div>A bit longer stor=
y:</div><div><br></div><div>Recently, when testing quicly=C2=A0[1] using a =
NIC with *hardware support* for UDP GSO offload, we noticed that it uses +3=
0% CPU cycles compared to TLS over TCP. This was due to our underlying cryp=
to not being optimized for short AEAD blocks, and due to the call overhead =
of doing header protection one packet=C2=A0at a time. [2]</div><div><br></d=
iv><div>We went on to implement our own AES-GCM library optimized for short=
 AEAD blocks (i.e. packet-sized), that also generates the header protection=
 vector while doing AEAD computation. By switching to the new crypto librar=
y, the performance drawback of the crypto itself compared to large AEAD blo=
cks when down from 30% to 10% [3].=C2=A0</div><div><br></div><div>The overh=
ead of header protection (relative to all the crypto cost) on the send side=
 is now between 0.3% to 1% when using packets sized around 1400 bytes to 15=
00 bytes. The ratio against the total transport cost is going to be less th=
an that, and in practice, I think that they are negligible. The story might=
 be a bit different on the receive-side due to the dependency between heade=
r protection and AEAD, though I am sure there would also be ways to optimiz=
e.</div><div><br></div><div>In terms of CPU cycles spent by the QUIC stack,=
 the overhead compared to TLS-over-TCP came down to 5%. quicly can now sust=
ain the send rate of 5Gbps while spending about 30% of CPU cycles of one mo=
dern Intel CPU core [2].</div><div><br></div><div>These numbers are observa=
tions from one TLS / QUIC stack doing micro-benchmarks, though I tend to th=
ink that sharing performance numbers lead to people having better expectati=
ons regarding the CPU cost of QUIC. Some of us were also concerned about th=
e cost of header protection (a.k.a. packet number encryption) [4]. Hence po=
sting here.<br clear=3D"all"><div><br></div><div>[1] <a href=3D"https://git=
hub.com/h2o/quicly">https://github.com/h2o/quicly</a></div><div><span style=
=3D"color:rgb(0,0,0)">[2]=C2=A0<a href=3D"https://github.com/h2o/quicly/pul=
l/359">https://github.com/h2o/quicly/pull/359</a></span></div><div><span st=
yle=3D"color:rgb(0,0,0)"></span>[3]=C2=A0<a href=3D"https://github.com/h2o/=
picotls/pull/310">https://github.com/h2o/picotls/pull/310</a></div><div>[4]=
=C2=A0<a href=3D"https://mailarchive.ietf.org/arch/msg/quic/Wy8JXwKyeYgWgRW=
izWDbyOLg0xc/">https://mailarchive.ietf.org/arch/msg/quic/Wy8JXwKyeYgWgRWiz=
WDbyOLg0xc/</a></div><div><br></div>-- <br><div dir=3D"ltr" class=3D"gmail_=
signature">Kazuho Oku</div></div></div></div></div></div></div>

--0000000000004fb7db05a8577f73--


From nobody Thu Jun 18 07:52:07 2020
Return-Path: <nibanks@microsoft.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C7AA03A12BE for <quic@ietfa.amsl.com>; Thu, 18 Jun 2020 07:52:05 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2
X-Spam-Level: 
X-Spam-Status: No, score=-2 tagged_above=-999 required=5 tests=[BAYES_00=-1.9,  DKIMWL_WL_HIGH=-0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, HTML_MESSAGE=0.001, HTTPS_HTTP_MISMATCH=0.1, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=microsoft.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id z9_sjQ80Ko4s for <quic@ietfa.amsl.com>; Thu, 18 Jun 2020 07:52:03 -0700 (PDT)
Received: from NAM06-BL2-obe.outbound.protection.outlook.com (mail-eopbgr650095.outbound.protection.outlook.com [40.107.65.95]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 45F253A12B8 for <quic@ietf.org>; Thu, 18 Jun 2020 07:52:01 -0700 (PDT)
ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=a1qItXel8qsTQRhbozNQRn6jK/s40/JdxgD1HkbLqBr9H0ocbaUttPx2X0AYZJXz9vCd83ViHquLBPgtGGSf63zh87zCG3BrHjWpQldJobVCWdFphxnJcNpDaDeu4qATSixAkOxaxZb0mTWg79A7AP/0k6mM1qXFxl6BbgJDIBt9B7JEq1ts9OfEUymx1fSAzYXhSFGeLuIlBlivpO2uavd8tw/F5+POYMHXxCI4wbs2DE/xACpoyBiPovUZZ1m7jv4pBMI5f8qmYSmZp+O4Dj2SBlI8ZI9PhVL8MdtMFk9Z2L9dhI7k5Latd79onNzUI+ZRQSULGz5QMK5ZteOxGg==
ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com;  s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=bACXSJIpq4Rqsk1637yyi4ZSiW+SHim577yXZ6nmY9w=; b=FFY9owevMCtttusE6RlatYiTjQevRTp47Aou+UoW1/FAGoXMXCyNIgaJYaxdLtNuxiHSB3CR2PPLCNd7n/faGfaKFqdRw6SgbIQ+LndExo/MDoHVhA0akjUQiI9bLRtDE8igUpyH3AaRuLo+EvldcgL7pyeNrV6rRuzMkx4qMctSacfifdj238UUdf/8X2vH0OAwECsB1lEMWIJcGQA14VxdBlHbT9gJJAQ65mHg7DHuvDZClpnByYMHhWc4b2K7T70N2NS2ifc/jMCis1yu7UFvK5aBJgG4A/GVYBwf911gNR2gUCmbyaXokauL3Y6YZUw8AO2b8l6UgtbjhC4WjQ==
ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=microsoft.com; dmarc=pass action=none header.from=microsoft.com; dkim=pass header.d=microsoft.com; arc=none
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=selector2; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=bACXSJIpq4Rqsk1637yyi4ZSiW+SHim577yXZ6nmY9w=; b=RpNYnK9W4cr8LC0bh9J+lXky4b6sFBnnIU5U11ANp5FjiarRYXOjvKSuedMgY/04W7X7X0ycuP+TIeOnscu9XPU7ms1fNkL3DNF20PTby3jOCTfHiRsyBjCXD0WLokakAaTCvt27m5aynvdd9poD8x5/R/wdXsr9ZGWhoMiShaQ=
Received: from CY1PR00MB0170.namprd00.prod.outlook.com (2a01:111:e400:c638::26) by CY1PR00MB0075.namprd00.prod.outlook.com (2a01:111:e400:c619::25) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.3153.0; Thu, 18 Jun 2020 14:51:58 +0000
Received: from CY1PR00MB0170.namprd00.prod.outlook.com ([fe80::7d0f:146e:c697:6989]) by CY1PR00MB0170.namprd00.prod.outlook.com ([fe80::7d0f:146e:c697:6989%3]) with mapi id 15.20.3153.000; Thu, 18 Jun 2020 14:51:58 +0000
From: Nick Banks <nibanks@microsoft.com>
To: Kazuho Oku <kazuhooku@gmail.com>, IETF QUIC WG <quic@ietf.org>
Subject: RE: AEAD and header encryption overhead in QUIC
Thread-Topic: AEAD and header encryption overhead in QUIC
Thread-Index: AQHWRUnYDE4fg8npukO25Wg8NKLEMqjecdOs
Date: Thu, 18 Jun 2020 14:51:58 +0000
Message-ID: <SN2PR00MB0173E37EEAD6CB56F7425DF4B39B0@SN2PR00MB0173.namprd00.prod.outlook.com>
References: <CANatvzz8F1H=DXMkBEhmKHnYM-HVG48TS9KwY=OP881Txkcodw@mail.gmail.com>
In-Reply-To: <CANatvzz8F1H=DXMkBEhmKHnYM-HVG48TS9KwY=OP881Txkcodw@mail.gmail.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
msip_labels: MSIP_Label_f42aa342-8706-4288-bd11-ebb85995028c_Enabled=True; MSIP_Label_f42aa342-8706-4288-bd11-ebb85995028c_SiteId=72f988bf-86f1-41af-91ab-2d7cd011db47; MSIP_Label_f42aa342-8706-4288-bd11-ebb85995028c_SetDate=2020-06-18T14:38:58.3938238Z; MSIP_Label_f42aa342-8706-4288-bd11-ebb85995028c_ContentBits=0; MSIP_Label_f42aa342-8706-4288-bd11-ebb85995028c_Method=Privileged
authentication-results: gmail.com; dkim=none (message not signed) header.d=none;gmail.com; dmarc=none action=none header.from=microsoft.com;
x-originating-ip: [66.235.1.136]
x-ms-publictraffictype: Email
x-ms-office365-filtering-ht: Tenant
x-ms-office365-filtering-correlation-id: 126713e8-c33e-4d08-1542-08d8139726b6
x-ms-traffictypediagnostic: CY1PR00MB0075:
x-microsoft-antispam-prvs: <CY1PR00MB0075F38101D793CB162F51D6B39B0@CY1PR00MB0075.namprd00.prod.outlook.com>
x-ms-oob-tlc-oobclassifiers: OLM:10000;
x-forefront-prvs: 0438F90F17
x-ms-exchange-senderadcheck: 1
x-microsoft-antispam: BCL:0;
x-microsoft-antispam-message-info: 0Fi/sRfbtGeX8rditubiMykZEiOWnhhuA9do5mktMKBq1IAqmdMTIHw2g4vuT4j5no1em6AvyGuwLlEUrvhvii7J8CNp9nbuADjJ+oaS/bQpy20PX22LaIr6hkgGLQI4G/t457XPM9ab8szHj0WwkUblkAhJy8wxeed8XehfSMraoNm1g5mdaFDKVxcEyigam/S3JfPVlzNtCPpgBgrCRYGAy5h/8pcLE/bhzMcuUAEKBBtSzSN7ZzLLQmmdL5jmrIRYhH4ME7xBDCh9/7dsxs1XZxfbLt82K2Rzh39UB6GyTW3ofLGDnbeD40ekW26QRFKVTJ+JCYAenIBVAvDKEP5ME1EHAUpNsvqWS/5/W0QtNexrIhePHIXF/AorK+U6tcDEYoXjEM/8c+fVj+H55Q==
x-forefront-antispam-report: CIP:255.255.255.255; CTRY:; LANG:en; SCL:1; SRV:;  IPV:NLI; SFV:NSPM; H:CY1PR00MB0170.namprd00.prod.outlook.com; PTR:; CAT:NONE;  SFTY:; SFS:(4636009)(376002)(366004)(346002)(136003)(396003)(39860400002)(76116006)(66946007)(91956017)(6512007)(966005)(33656002)(86362001)(110136005)(316002)(82950400001)(82960400001)(52536014)(64756008)(66446008)(478600001)(66476007)(5660300002)(66556008)(9686003)(2906002)(71200400001)(8936002)(186003)(26005)(9326002)(83380400001)(6486002)(53546011)(8990500004)(8676002)(166002)(10290500003)(6506007); DIR:OUT; SFP:1102; 
x-ms-exchange-antispam-messagedata: N4eQZQZMzlyI9AOZdurAyaLwEeSpFeO+hf/5yonDNGi/AFiX7OFkU5Su7XVWdwZWxJ306k7c7z0IhuviS4BHZKZWCPyDhD6FMqunkvA+v0xswv2lc2nJZkbGPKG1GpAh4g2kJcqJ8GVitDoBYetu3rP5Gaa06zvUBmxEBqa00oyQz0Oi5TLiouclEfg5Rqk2zlBg1cjKur3Cz0DWf5JDnyQBcz3c2iZ0C4B5RYciOutqSpyV1+95JopPUIwm9vd20dEuT2PumdrglXt5wRgyHyU4LHbFhXKelzaOqsFJJ++9QuewHbjR/vEsbGHWDcbZQriY5QJQytb87nUCZsET9XEf6qWggIAkW4IdmsKpE9cs/n4HNGzyrNz1k1rpbdMawR6bmDhxc9Hsjjg1gvAMiMeu3kuAh/30YuHZsXCsvh4E7nip9/2owdrFanlt9iYBJHIUKxlYXZNW0OMd3jEi6XZFauGmVJeUBj7B/z9a37I=
x-ms-exchange-transport-forked: True
Content-Type: multipart/alternative; boundary="_000_SN2PR00MB0173E37EEAD6CB56F7425DF4B39B0SN2PR00MB0173namp_"
MIME-Version: 1.0
X-OriginatorOrg: microsoft.com
X-MS-Exchange-CrossTenant-AuthAs: Internal
X-MS-Exchange-CrossTenant-AuthSource: CY1PR00MB0170.namprd00.prod.outlook.com
X-MS-Exchange-CrossTenant-Network-Message-Id: 126713e8-c33e-4d08-1542-08d8139726b6
X-MS-Exchange-CrossTenant-originalarrivaltime: 18 Jun 2020 14:51:58.1902 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Hosted
X-MS-Exchange-CrossTenant-id: 72f988bf-86f1-41af-91ab-2d7cd011db47
X-MS-Exchange-CrossTenant-mailboxtype: HOSTED
X-MS-Exchange-CrossTenant-userprincipalname: PDygj1kSQe89696yWSEpQf67BytCubU9q/UZJYquRX90mevWEuerezVlM+8JUox8qlMDDt4rvV1Dtn3Yzu46/A==
X-MS-Exchange-Transport-CrossTenantHeadersStamped: CY1PR00MB0075
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/sTEmwiK-NY9dlgKkMus46AYceZs>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 18 Jun 2020 14:52:06 -0000

--_000_SN2PR00MB0173E37EEAD6CB56F7425DF4B39B0SN2PR00MB0173namp_
Content-Type: text/plain; charset="Windows-1252"
Content-Transfer-Encoding: quoted-printable

Hi Kazuho,

Thanks for the information! It=92s very interesting. You might try this too=
:

We=92ve found that an easy way to minimize the CPU cost of the header prote=
ction (on send and receive) is batching. If you copy 8 packet headers into =
a single contiguous block of memory, you can do a single crypto operation o=
n it. The cost of doing one batch of 8 is essentially the same as doing jus=
t a single header. This effectively cuts the cost of header protection to 1=
/8th the original cost. Obviously this only works if you have enough packet=
s to batch process, but for high throughput tests, you usually do.

Feel free to take a look at the MsQuic code (receive path<https://github.co=
m/microsoft/msquic/blob/master/src/core/connection.c#L4784>, send path<http=
s://github.com/microsoft/msquic/blob/master/src/core/packet_builder.c#L728>=
). The nice thing about this design (for those of us who don=92t have as mu=
ch special crypto expertise) is that it doesn=92t require any special crypt=
o changes. I=92d be interested to see if our two approaches could be combin=
ed for even better performance!

Thanks,
- Nick

From: Kazuho Oku<mailto:kazuhooku@gmail.com>
Sent: Thursday, June 18, 2020 1:24 AM
To: IETF QUIC WG<mailto:quic@ietf.org>
Subject: AEAD and header encryption overhead in QUIC

Hello folks,

TL;DR: When existing crypto libraries are used, the AEAD cost of QUIC can b=
ecome somewhat greater than that of TLS over TCP, due to those crypto libra=
ries not optimized for handling short AEAD blocks. Crypto libraries optimiz=
ed for short AEAD blocks can tighten the gap. Also, crypto libraries optimi=
zed for QUIC can combine AEAD operation and header protection vector genera=
tion, effectively eliminating the cost of header protection.

A bit longer story:

Recently, when testing quicly [1] using a NIC with *hardware support* for U=
DP GSO offload, we noticed that it uses +30% CPU cycles compared to TLS ove=
r TCP. This was due to our underlying crypto not being optimized for short =
AEAD blocks, and due to the call overhead of doing header protection one pa=
cket at a time. [2]

We went on to implement our own AES-GCM library optimized for short AEAD bl=
ocks (i.e. packet-sized), that also generates the header protection vector =
while doing AEAD computation. By switching to the new crypto library, the p=
erformance drawback of the crypto itself compared to large AEAD blocks when=
 down from 30% to 10% [3].

The overhead of header protection (relative to all the crypto cost) on the =
send side is now between 0.3% to 1% when using packets sized around 1400 by=
tes to 1500 bytes. The ratio against the total transport cost is going to b=
e less than that, and in practice, I think that they are negligible. The st=
ory might be a bit different on the receive-side due to the dependency betw=
een header protection and AEAD, though I am sure there would also be ways t=
o optimize.

In terms of CPU cycles spent by the QUIC stack, the overhead compared to TL=
S-over-TCP came down to 5%. quicly can now sustain the send rate of 5Gbps w=
hile spending about 30% of CPU cycles of one modern Intel CPU core [2].

These numbers are observations from one TLS / QUIC stack doing micro-benchm=
arks, though I tend to think that sharing performance numbers lead to peopl=
e having better expectations regarding the CPU cost of QUIC. Some of us wer=
e also concerned about the cost of header protection (a.k.a. packet number =
encryption) [4]. Hence posting here.

[1] https://github.com/h2o/quicly<https://nam06.safelinks.protection.outloo=
k.com/?url=3Dhttps%3A%2F%2Fgithub.com%2Fh2o%2Fquicly&data=3D02%7C01%7Cniban=
ks%40microsoft.com%7Cce44384090cd41b9874f08d81360f9dc%7C72f988bf86f141af91a=
b2d7cd011db47%7C1%7C0%7C637280654511448872&sdata=3DhjeugkkQwFtQNx7UriHYA9eQ=
3z1z9omSxB6XRPiUF1Y%3D&reserved=3D0>
[2] https://github.com/h2o/quicly/pull/359<https://nam06.safelinks.protecti=
on.outlook.com/?url=3Dhttps%3A%2F%2Fgithub.com%2Fh2o%2Fquicly%2Fpull%2F359&=
data=3D02%7C01%7Cnibanks%40microsoft.com%7Cce44384090cd41b9874f08d81360f9dc=
%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637280654511458858&sdata=3DVS=
e%2Br2rp81U7RIFhRaJ%2FAh4YGXobO4F90v7lUcxqMoA%3D&reserved=3D0>
[3] https://github.com/h2o/picotls/pull/310<https://nam06.safelinks.protect=
ion.outlook.com/?url=3Dhttps%3A%2F%2Fgithub.com%2Fh2o%2Fpicotls%2Fpull%2F31=
0&data=3D02%7C01%7Cnibanks%40microsoft.com%7Cce44384090cd41b9874f08d81360f9=
dc%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637280654511458858&sdata=3D=
xbHdVd9cewW%2BlwSlIQtG4f3sTGYFxqM6c60qnVFkVbI%3D&reserved=3D0>
[4] https://mailarchive.ietf.org/arch/msg/quic/Wy8JXwKyeYgWgRWizWDbyOLg0xc/=
<https://nam06.safelinks.protection.outlook.com/?url=3Dhttps%3A%2F%2Fmailar=
chive.ietf.org%2Farch%2Fmsg%2Fquic%2FWy8JXwKyeYgWgRWizWDbyOLg0xc%2F&data=3D=
02%7C01%7Cnibanks%40microsoft.com%7Cce44384090cd41b9874f08d81360f9dc%7C72f9=
88bf86f141af91ab2d7cd011db47%7C1%7C0%7C637280654511468853&sdata=3DZT3%2FAtl=
8HpbN5KKYdI0vY8cmVnoeSJaS1Jys%2F6QiQ9A%3D&reserved=3D0>

--
Kazuho Oku


--_000_SN2PR00MB0173E37EEAD6CB56F7425DF4B39B0SN2PR00MB0173namp_
Content-Type: text/html; charset="Windows-1252"
Content-Transfer-Encoding: quoted-printable

<html xmlns:o=3D"urn:schemas-microsoft-com:office:office" xmlns:w=3D"urn:sc=
hemas-microsoft-com:office:word" xmlns:m=3D"http://schemas.microsoft.com/of=
fice/2004/12/omml" xmlns=3D"http://www.w3.org/TR/REC-html40">
<head>
<meta http-equiv=3D"Content-Type" content=3D"text/html; charset=3DWindows-1=
252">
<meta name=3D"Generator" content=3D"Microsoft Word 15 (filtered medium)">
<style><!--
/* Font Definitions */
@font-face
	{font-family:"Cambria Math";
	panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
	{font-family:Calibri;
	panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
	{margin:0in;
	margin-bottom:.0001pt;
	font-size:11.0pt;
	font-family:"Calibri",sans-serif;}
a:link, span.MsoHyperlink
	{mso-style-priority:99;
	color:blue;
	text-decoration:underline;}
.MsoChpDefault
	{mso-style-type:export-only;}
@page WordSection1
	{size:8.5in 11.0in;
	margin:1.0in 1.0in 1.0in 1.0in;}
div.WordSection1
	{page:WordSection1;}
--></style>
</head>
<body lang=3D"EN-US" link=3D"blue" vlink=3D"#954F72">
<div class=3D"WordSection1">
<p class=3D"MsoNormal">Hi Kazuho,</p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">Thanks for the information! It=92s very interesting.=
 You might try this too:</p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">We=92ve found that an easy way to minimize the CPU c=
ost of the header protection (on send and receive) is batching. If you copy=
 8 packet headers into a single contiguous block of memory, you can do a si=
ngle crypto operation on it. The cost
 of doing one batch of 8 is essentially the same as doing just a single hea=
der. This effectively cuts the cost of header protection to 1/8<sup>th</sup=
> the original cost. Obviously this only works if you have enough packets t=
o batch process, but for high throughput
 tests, you usually do.</p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">Feel free to take a look at the MsQuic code (<a href=
=3D"https://github.com/microsoft/msquic/blob/master/src/core/connection.c#L=
4784">receive path</a>,
<a href=3D"https://github.com/microsoft/msquic/blob/master/src/core/packet_=
builder.c#L728">
send path</a>). The nice thing about this design (for those of us who don=
=92t have as much special crypto expertise) is that it doesn=92t require an=
y special crypto changes. I=92d be interested to see if our two approaches =
could be combined for even better performance!</p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">Thanks,</p>
<p class=3D"MsoNormal">- Nick</p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<div style=3D"mso-element:para-border-div;border:none;border-top:solid #E1E=
1E1 1.0pt;padding:3.0pt 0in 0in 0in">
<p class=3D"MsoNormal" style=3D"border:none;padding:0in"><b>From: </b><a hr=
ef=3D"mailto:kazuhooku@gmail.com">Kazuho Oku</a><br>
<b>Sent: </b>Thursday, June 18, 2020 1:24 AM<br>
<b>To: </b><a href=3D"mailto:quic@ietf.org">IETF QUIC WG</a><br>
<b>Subject: </b>AEAD and header encryption overhead in QUIC</p>
</div>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<div>
<div>
<div>
<div>
<div>
<p class=3D"MsoNormal">Hello folks,<o:p></o:p></p>
<div>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
</div>
<div>
<p class=3D"MsoNormal">TL;DR: When existing crypto libraries are used, the =
AEAD cost of QUIC can become somewhat greater than that of TLS over TCP, du=
e to those crypto libraries not optimized for handling short AEAD blocks. C=
rypto libraries optimized for&nbsp;short
 AEAD blocks can tighten the gap. Also, crypto libraries optimized for QUIC=
 can combine AEAD operation and header protection vector generation, effect=
ively eliminating the cost of header protection.<o:p></o:p></p>
</div>
<div>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
</div>
<div>
<p class=3D"MsoNormal">A bit longer story:<o:p></o:p></p>
</div>
<div>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
</div>
<div>
<p class=3D"MsoNormal">Recently, when testing quicly&nbsp;[1] using a NIC w=
ith *hardware support* for UDP GSO offload, we noticed that it uses &#43;30=
% CPU cycles compared to TLS over TCP. This was due to our underlying crypt=
o not being optimized for short AEAD blocks,
 and due to the call overhead of doing header protection one packet&nbsp;at=
 a time. [2]<o:p></o:p></p>
</div>
<div>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
</div>
<div>
<p class=3D"MsoNormal">We went on to implement our own AES-GCM library opti=
mized for short AEAD blocks (i.e. packet-sized), that also generates the he=
ader protection vector while doing AEAD computation. By switching to the ne=
w crypto library, the performance
 drawback of the crypto itself compared to large AEAD blocks when down from=
 30% to 10% [3].&nbsp;<o:p></o:p></p>
</div>
<div>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
</div>
<div>
<p class=3D"MsoNormal">The overhead of header protection (relative to all t=
he crypto cost) on the send side is now between 0.3% to 1% when using packe=
ts sized around 1400 bytes to 1500 bytes. The ratio against the total trans=
port cost is going to be less than
 that, and in practice, I think that they are negligible. The story might b=
e a bit different on the receive-side due to the dependency between header =
protection and AEAD, though I am sure there would also be ways to optimize.=
<o:p></o:p></p>
</div>
<div>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
</div>
<div>
<p class=3D"MsoNormal">In terms of CPU cycles spent by the QUIC stack, the =
overhead compared to TLS-over-TCP came down to 5%. quicly can now sustain t=
he send rate of 5Gbps while spending about 30% of CPU cycles of one modern =
Intel CPU core [2].<o:p></o:p></p>
</div>
<div>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
</div>
<div>
<p class=3D"MsoNormal">These numbers are observations from one TLS / QUIC s=
tack doing micro-benchmarks, though I tend to think that sharing performanc=
e numbers lead to people having better expectations regarding the CPU cost =
of QUIC. Some of us were also concerned
 about the cost of header protection (a.k.a. packet number encryption) [4].=
 Hence posting here.<br clear=3D"all">
<o:p></o:p></p>
<div>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
</div>
<div>
<p class=3D"MsoNormal">[1] <a href=3D"https://nam06.safelinks.protection.ou=
tlook.com/?url=3Dhttps%3A%2F%2Fgithub.com%2Fh2o%2Fquicly&amp;data=3D02%7C01=
%7Cnibanks%40microsoft.com%7Cce44384090cd41b9874f08d81360f9dc%7C72f988bf86f=
141af91ab2d7cd011db47%7C1%7C0%7C637280654511448872&amp;sdata=3DhjeugkkQwFtQ=
Nx7UriHYA9eQ3z1z9omSxB6XRPiUF1Y%3D&amp;reserved=3D0">
https://github.com/h2o/quicly</a><o:p></o:p></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"color:black">[2]&nbsp;<a href=3D"http=
s://nam06.safelinks.protection.outlook.com/?url=3Dhttps%3A%2F%2Fgithub.com%=
2Fh2o%2Fquicly%2Fpull%2F359&amp;data=3D02%7C01%7Cnibanks%40microsoft.com%7C=
ce44384090cd41b9874f08d81360f9dc%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0=
%7C637280654511458858&amp;sdata=3DVSe%2Br2rp81U7RIFhRaJ%2FAh4YGXobO4F90v7lU=
cxqMoA%3D&amp;reserved=3D0">https://github.com/h2o/quicly/pull/359</a></spa=
n><o:p></o:p></p>
</div>
<div>
<p class=3D"MsoNormal">[3]&nbsp;<a href=3D"https://nam06.safelinks.protecti=
on.outlook.com/?url=3Dhttps%3A%2F%2Fgithub.com%2Fh2o%2Fpicotls%2Fpull%2F310=
&amp;data=3D02%7C01%7Cnibanks%40microsoft.com%7Cce44384090cd41b9874f08d8136=
0f9dc%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637280654511458858&amp;s=
data=3DxbHdVd9cewW%2BlwSlIQtG4f3sTGYFxqM6c60qnVFkVbI%3D&amp;reserved=3D0">h=
ttps://github.com/h2o/picotls/pull/310</a><o:p></o:p></p>
</div>
<div>
<p class=3D"MsoNormal">[4]&nbsp;<a href=3D"https://nam06.safelinks.protecti=
on.outlook.com/?url=3Dhttps%3A%2F%2Fmailarchive.ietf.org%2Farch%2Fmsg%2Fqui=
c%2FWy8JXwKyeYgWgRWizWDbyOLg0xc%2F&amp;data=3D02%7C01%7Cnibanks%40microsoft=
.com%7Cce44384090cd41b9874f08d81360f9dc%7C72f988bf86f141af91ab2d7cd011db47%=
7C1%7C0%7C637280654511468853&amp;sdata=3DZT3%2FAtl8HpbN5KKYdI0vY8cmVnoeSJaS=
1Jys%2F6QiQ9A%3D&amp;reserved=3D0">https://mailarchive.ietf.org/arch/msg/qu=
ic/Wy8JXwKyeYgWgRWizWDbyOLg0xc/</a><o:p></o:p></p>
</div>
<div>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
</div>
<p class=3D"MsoNormal">-- <o:p></o:p></p>
</div>
</div>
</div>
</div>
</div>
</div>
<p class=3D"MsoNormal">Kazuho Oku<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
</div>
</body>
</html>

--_000_SN2PR00MB0173E37EEAD6CB56F7425DF4B39B0SN2PR00MB0173namp_--


From nobody Thu Jun 18 10:21:27 2020
Return-Path: <matt.joras@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 9E7843A0CE1 for <quic@ietfa.amsl.com>; Thu, 18 Jun 2020 10:21:25 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id vMPfIgtXYKtw for <quic@ietfa.amsl.com>; Thu, 18 Jun 2020 10:21:24 -0700 (PDT)
Received: from mail-vk1-xa2e.google.com (mail-vk1-xa2e.google.com [IPv6:2607:f8b0:4864:20::a2e]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id C776E3A0CA7 for <quic@ietf.org>; Thu, 18 Jun 2020 10:21:23 -0700 (PDT)
Received: by mail-vk1-xa2e.google.com with SMTP id q69so1619512vkq.10 for <quic@ietf.org>; Thu, 18 Jun 2020 10:21:23 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=eifcHAi/b6anFfzog/tzOoN8zc2F9SgD3JnvDRk6/NM=; b=UMs0+EkZii4y16k6vYl9lBdBDLKmGSnk4cR3JTVdSh83ugeABj0iWLGGcyVyAxrhsh rlBDqk5PNukWBddaU4T/ptRcxO0ba/EZhuzh+p3zgibOdCubzCdpoy4oOiWgbVmIC0Uf raOizABY42/H/aMcM7XnIVdnn6LuBiu2Q8ASM4Lpdgo4mwN08jahGZ+ysTVHR7jSmR2Y cyp7fIiEiKaTXuy8vtcyVImNd0M3I4N3Fn3sel42j1qJmHoUhUA33g/gl90KgG/Ga9in oYJST552Vtuc/odROP+Sx0AcsV4tl76gsHxZdfGER6EW4wIAPMHkqdIdO7ywQ3pFdszM MziA==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=eifcHAi/b6anFfzog/tzOoN8zc2F9SgD3JnvDRk6/NM=; b=TOFgKxCUIOqxEfz1u6bhhOPn63pIfvqpiWQRc4xrhMVXCTyoyMXpjwNHYchAtuDMmn YCbPDq8jhNY8hGIjIzl9W0EU4t8TmhsWhfvxEunxttPQLqqnpJktp2sXukzK9DCUEmq9 +C+vopVamcI8s+k6lfKWKWebptvTQcyFluowTDV8FoqxHlWf4eRlA2sTnLV0sEJ2Rmw0 DqlQYAgTl8ldqUPqHjZCaBGRig6pQjDXUF+tuoFMm8Hwp79faJHx89XN0KWvQOj6cynb /spNcOaXUTBoFlPGsTzJx52igtxaXfoWKMJopQyWDNbpvBYZUOwDiHsLmZ9V95iEsH6Q KLDg==
X-Gm-Message-State: AOAM5331Gz2wY8BuyH3OsbcYerFPIqnPDj230w+gT4DFgDTJ77g8kTya saGktQZAhCuaSjZF/Sd+IQbsTHZN7aT7NMhQ8OE=
X-Google-Smtp-Source: ABdhPJx/XygiAPNA3v8PfqlKZ5mEEcYb0GezgRYp7rFJcjyeY1dPc+vk3+eu2BurlML5kz/AvysNcZsc6SIOQwKyPEw=
X-Received: by 2002:a1f:9511:: with SMTP id x17mr4391054vkd.101.1592500882717;  Thu, 18 Jun 2020 10:21:22 -0700 (PDT)
MIME-Version: 1.0
References: <CANatvzz8F1H=DXMkBEhmKHnYM-HVG48TS9KwY=OP881Txkcodw@mail.gmail.com>
In-Reply-To: <CANatvzz8F1H=DXMkBEhmKHnYM-HVG48TS9KwY=OP881Txkcodw@mail.gmail.com>
From: Matt Joras <matt.joras@gmail.com>
Date: Thu, 18 Jun 2020 10:21:11 -0700
Message-ID: <CADdTf+i+LZ98GgNhFNVcuoVczC=jCQE-TqWbCqhrpR7=Z2knWg@mail.gmail.com>
Subject: Re: AEAD and header encryption overhead in QUIC
To: Kazuho Oku <kazuhooku@gmail.com>
Cc: IETF QUIC WG <quic@ietf.org>
Content-Type: multipart/alternative; boundary="0000000000006b402405a85f01b4"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/lbqb67gA34Cj0o_8fmeXNzlcayc>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 18 Jun 2020 17:21:26 -0000

--0000000000006b402405a85f01b4
Content-Type: text/plain; charset="UTF-8"

Wow, really interesting results, thanks for sharing!

I was curious, since it's not mentioned on the PRs that I saw, were these
tests done with any adjustments to the ACK frequency on the client? Or is
this with the default ACK policy of ACKing every other?

Matt Joras

On Thu, Jun 18, 2020 at 1:24 AM Kazuho Oku <kazuhooku@gmail.com> wrote:

> Hello folks,
>
> TL;DR: When existing crypto libraries are used, the AEAD cost of QUIC can
> become somewhat greater than that of TLS over TCP, due to those crypto
> libraries not optimized for handling short AEAD blocks. Crypto libraries
> optimized for short AEAD blocks can tighten the gap. Also, crypto libraries
> optimized for QUIC can combine AEAD operation and header protection vector
> generation, effectively eliminating the cost of header protection.
>
> A bit longer story:
>
> Recently, when testing quicly [1] using a NIC with *hardware support* for
> UDP GSO offload, we noticed that it uses +30% CPU cycles compared to TLS
> over TCP. This was due to our underlying crypto not being optimized for
> short AEAD blocks, and due to the call overhead of doing header protection
> one packet at a time. [2]
>
> We went on to implement our own AES-GCM library optimized for short AEAD
> blocks (i.e. packet-sized), that also generates the header protection
> vector while doing AEAD computation. By switching to the new crypto
> library, the performance drawback of the crypto itself compared to large
> AEAD blocks when down from 30% to 10% [3].
>
> The overhead of header protection (relative to all the crypto cost) on the
> send side is now between 0.3% to 1% when using packets sized around 1400
> bytes to 1500 bytes. The ratio against the total transport cost is going to
> be less than that, and in practice, I think that they are negligible. The
> story might be a bit different on the receive-side due to the dependency
> between header protection and AEAD, though I am sure there would also be
> ways to optimize.
>
> In terms of CPU cycles spent by the QUIC stack, the overhead compared to
> TLS-over-TCP came down to 5%. quicly can now sustain the send rate of 5Gbps
> while spending about 30% of CPU cycles of one modern Intel CPU core [2].
>
> These numbers are observations from one TLS / QUIC stack doing
> micro-benchmarks, though I tend to think that sharing performance numbers
> lead to people having better expectations regarding the CPU cost of QUIC.
> Some of us were also concerned about the cost of header protection (a.k.a.
> packet number encryption) [4]. Hence posting here.
>
> [1] https://github.com/h2o/quicly
> [2] https://github.com/h2o/quicly/pull/359
> [3] https://github.com/h2o/picotls/pull/310
> [4]
> https://mailarchive.ietf.org/arch/msg/quic/Wy8JXwKyeYgWgRWizWDbyOLg0xc/
>
> --
> Kazuho Oku
>

--0000000000006b402405a85f01b4
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div><div><div>Wow, really interesting results, thanks for=
 sharing!<br><br></div>I was curious, since it&#39;s not mentioned on the P=
Rs that I saw, were these tests done with any adjustments to the ACK freque=
ncy on the client? Or is this with the default ACK policy of ACKing every o=
ther?<br><br></div></div>Matt Joras<br></div><br><div class=3D"gmail_quote"=
><div dir=3D"ltr" class=3D"gmail_attr">On Thu, Jun 18, 2020 at 1:24 AM Kazu=
ho Oku &lt;<a href=3D"mailto:kazuhooku@gmail.com">kazuhooku@gmail.com</a>&g=
t; wrote:<br></div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0p=
x 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div d=
ir=3D"ltr"><div dir=3D"ltr"><div dir=3D"ltr"><div dir=3D"ltr"><div dir=3D"l=
tr">Hello folks,<div><br></div><div>TL;DR: When existing crypto libraries a=
re used, the AEAD cost of QUIC can become somewhat greater than that of TLS=
 over TCP, due to those crypto libraries not optimized for handling short A=
EAD blocks. Crypto libraries optimized for=C2=A0short AEAD blocks can tight=
en the gap. Also, crypto libraries optimized for QUIC can combine AEAD oper=
ation and header protection vector generation, effectively eliminating the =
cost of header protection.</div><div><br></div><div>A bit longer story:</di=
v><div><br></div><div>Recently, when testing quicly=C2=A0[1] using a NIC wi=
th *hardware support* for UDP GSO offload, we noticed that it uses +30% CPU=
 cycles compared to TLS over TCP. This was due to our underlying crypto not=
 being optimized for short AEAD blocks, and due to the call overhead of doi=
ng header protection one packet=C2=A0at a time. [2]</div><div><br></div><di=
v>We went on to implement our own AES-GCM library optimized for short AEAD =
blocks (i.e. packet-sized), that also generates the header protection vecto=
r while doing AEAD computation. By switching to the new crypto library, the=
 performance drawback of the crypto itself compared to large AEAD blocks wh=
en down from 30% to 10% [3].=C2=A0</div><div><br></div><div>The overhead of=
 header protection (relative to all the crypto cost) on the send side is no=
w between 0.3% to 1% when using packets sized around 1400 bytes to 1500 byt=
es. The ratio against the total transport cost is going to be less than tha=
t, and in practice, I think that they are negligible. The story might be a =
bit different on the receive-side due to the dependency between header prot=
ection and AEAD, though I am sure there would also be ways to optimize.</di=
v><div><br></div><div>In terms of CPU cycles spent by the QUIC stack, the o=
verhead compared to TLS-over-TCP came down to 5%. quicly can now sustain th=
e send rate of 5Gbps while spending about 30% of CPU cycles of one modern I=
ntel CPU core [2].</div><div><br></div><div>These numbers are observations =
from one TLS / QUIC stack doing micro-benchmarks, though I tend to think th=
at sharing performance numbers lead to people having better expectations re=
garding the CPU cost of QUIC. Some of us were also concerned about the cost=
 of header protection (a.k.a. packet number encryption) [4]. Hence posting =
here.<br clear=3D"all"><div><br></div><div>[1] <a href=3D"https://github.co=
m/h2o/quicly" target=3D"_blank">https://github.com/h2o/quicly</a></div><div=
><span style=3D"color:rgb(0,0,0)">[2]=C2=A0<a href=3D"https://github.com/h2=
o/quicly/pull/359" target=3D"_blank">https://github.com/h2o/quicly/pull/359=
</a></span></div><div><span style=3D"color:rgb(0,0,0)"></span>[3]=C2=A0<a h=
ref=3D"https://github.com/h2o/picotls/pull/310" target=3D"_blank">https://g=
ithub.com/h2o/picotls/pull/310</a></div><div>[4]=C2=A0<a href=3D"https://ma=
ilarchive.ietf.org/arch/msg/quic/Wy8JXwKyeYgWgRWizWDbyOLg0xc/" target=3D"_b=
lank">https://mailarchive.ietf.org/arch/msg/quic/Wy8JXwKyeYgWgRWizWDbyOLg0x=
c/</a></div><div><br></div>-- <br><div dir=3D"ltr">Kazuho Oku</div></div></=
div></div></div></div></div>
</blockquote></div>

--0000000000006b402405a85f01b4--


From nobody Thu Jun 18 10:35:57 2020
Return-Path: <mikkelfj@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 903583A0A8F for <quic@ietfa.amsl.com>; Thu, 18 Jun 2020 10:35:56 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.096
X-Spam-Level: 
X-Spam-Status: No, score=-1.096 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, FREEMAIL_REPLY=1, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, UNPARSEABLE_RELAY=0.001, URIBL_BLOCKED=0.001] autolearn=no autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id xVFlK4PrCpD0 for <quic@ietfa.amsl.com>; Thu, 18 Jun 2020 10:35:54 -0700 (PDT)
Received: from mail-yb1-xb29.google.com (mail-yb1-xb29.google.com [IPv6:2607:f8b0:4864:20::b29]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id B09F63A09F2 for <quic@ietf.org>; Thu, 18 Jun 2020 10:35:54 -0700 (PDT)
Received: by mail-yb1-xb29.google.com with SMTP id o4so3538069ybp.0 for <quic@ietf.org>; Thu, 18 Jun 2020 10:35:54 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=from:in-reply-to:references:mime-version:date:message-id:subject:to :cc; bh=mBNuN5B/fuaeZ6XT8QH1cy2XvzA+QdVSUay5dUf79II=; b=Ad9SgUOoTqQ6dkhRkx4DbXM9Y5ALAU0wRsT1POx1hcgvyB+0BkXclJmrxChN4o+RVV 3NvhkNgAIiW0w4kpKFwO7E3YDHKaFoPnAnOJ0b0zLd7v6ynhfqEQcDTbRt7HBtjj2hWe eRkaycrClVbjeh13kR3NrZlC1mu6D+RVBJsYLJZJBwuwaUkV3P/TGU0UHGJFm+mzsGFn Yld1Q31QWml0voLiCyoi+tavRViuq3CMIvxOzDvYjiDCP6ml4F7K+OE1NESclQhn5hFP 54Bk+cPdSuUfJD8LkWxPseku2Oa0EwaXWW7rAFYJp4m07BS8LuNqAqVyL3q8cEKTOpkz 3Eiw==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:in-reply-to:references:mime-version:date :message-id:subject:to:cc; bh=mBNuN5B/fuaeZ6XT8QH1cy2XvzA+QdVSUay5dUf79II=; b=exAOZAVGRvKXWFzr8VfOAXsCUFHRwmEhAw9Y21/QjuPgSvAQ7mlfqqQW9hqmjFMM9v ZNMVWibhoGtsxr/A186Pz8cHtQ9MufipeTsMmWfROcvNBQ7FX7SN5+1iGDMpocMebkT9 42i65+RFLhoBcZX5wtkkUFxaRB1GO4ZddfJRFn+yoriO4f/fOstz6aNcu+3evLFt7HJH iLrh2sX6438oOMJzTIBJY2xXTMZC9i1/V2/p7B8TY1YbaOTcfEzLvfjURxgdjQjpZ3u+ oeu8++2uga9qpQco1Mn2r3vYpeEP+H6mhydmbLd8Vp2s18EEIDU0BcxTg9BiMmDyQ+6l fo+w==
X-Gm-Message-State: AOAM532UksWcbgoOEbh/MYbOMQr/t4fBHi0hBxNIrF88ugtxtDIdVTPB ZiO2Zwvf6l9/AwkDZyMnah+hHWaRx2NBSSJiCb8=
X-Google-Smtp-Source: ABdhPJx9oQFflTCqO6QSqpMEkZ+bzQ4cJKXRExidKtSVKOfBOD70950DIUYkp58pMT+ajNBvHTOK9D76X6TngQ9Ikcs=
X-Received: by 2002:a25:31c6:: with SMTP id x189mr8506966ybx.264.1592501753793;  Thu, 18 Jun 2020 10:35:53 -0700 (PDT)
Received: from 1058052472880 named unknown by gmailapi.google.com with HTTPREST; Thu, 18 Jun 2020 10:35:53 -0700
From: =?UTF-8?Q?Mikkel_Fahn=C3=B8e_J=C3=B8rgensen?= <mikkelfj@gmail.com>
In-Reply-To: <CADdTf+i+LZ98GgNhFNVcuoVczC=jCQE-TqWbCqhrpR7=Z2knWg@mail.gmail.com>
References: <CANatvzz8F1H=DXMkBEhmKHnYM-HVG48TS9KwY=OP881Txkcodw@mail.gmail.com> <CADdTf+i+LZ98GgNhFNVcuoVczC=jCQE-TqWbCqhrpR7=Z2knWg@mail.gmail.com>
MIME-Version: 1.0
Date: Thu, 18 Jun 2020 10:35:53 -0700
Message-ID: <CAN1APdft3UU1dfKY_UxRaLy2xeCYSXQT3=k53=96OO_Gu1X_cw@mail.gmail.com>
Subject: Re: AEAD and header encryption overhead in QUIC
To: Kazuho Oku <kazuhooku@gmail.com>, Matt Joras <matt.joras@gmail.com>
Cc: IETF QUIC WG <quic@ietf.org>
Content-Type: multipart/alternative; boundary="00000000000056d08905a85f35d7"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/lj5R74oIrQHnbuzLMjJ-XFAuoUA>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 18 Jun 2020 17:35:57 -0000

--00000000000056d08905a85f35d7
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

I agree, thanks for sharing. I am not too surprised by the results though.

I am a bit surprised that it is possible to use bulk encrypt headers as
Nick suggests as I would have thought the key / nonce was associated with
the packet, but I haven=E2=80=99t looked closely recently.

Kind Regards,
Mikkel Fahn=C3=B8e J=C3=B8rgensen


On 18 June 2020 at 19.21.44, Matt Joras (matt.joras@gmail.com) wrote:

Wow, really interesting results, thanks for sharing!

I was curious, since it's not mentioned on the PRs that I saw, were these
tests done with any adjustments to the ACK frequency on the client? Or is
this with the default ACK policy of ACKing every other?

Matt Joras

On Thu, Jun 18, 2020 at 1:24 AM Kazuho Oku <kazuhooku@gmail.com> wrote:

> Hello folks,
>
> TL;DR: When existing crypto libraries are used, the AEAD cost of QUIC can
> become somewhat greater than that of TLS over TCP, due to those crypto
> libraries not optimized for handling short AEAD blocks. Crypto libraries
> optimized for short AEAD blocks can tighten the gap. Also, crypto librari=
es
> optimized for QUIC can combine AEAD operation and header protection vecto=
r
> generation, effectively eliminating the cost of header protection.
>
> A bit longer story:
>
> Recently, when testing quicly [1] using a NIC with *hardware support* for
> UDP GSO offload, we noticed that it uses +30% CPU cycles compared to TLS
> over TCP. This was due to our underlying crypto not being optimized for
> short AEAD blocks, and due to the call overhead of doing header protectio=
n
> one packet at a time. [2]
>
> We went on to implement our own AES-GCM library optimized for short AEAD
> blocks (i.e. packet-sized), that also generates the header protection
> vector while doing AEAD computation. By switching to the new crypto
> library, the performance drawback of the crypto itself compared to large
> AEAD blocks when down from 30% to 10% [3].
>
> The overhead of header protection (relative to all the crypto cost) on th=
e
> send side is now between 0.3% to 1% when using packets sized around 1400
> bytes to 1500 bytes. The ratio against the total transport cost is going =
to
> be less than that, and in practice, I think that they are negligible. The
> story might be a bit different on the receive-side due to the dependency
> between header protection and AEAD, though I am sure there would also be
> ways to optimize.
>
> In terms of CPU cycles spent by the QUIC stack, the overhead compared to
> TLS-over-TCP came down to 5%. quicly can now sustain the send rate of 5Gb=
ps
> while spending about 30% of CPU cycles of one modern Intel CPU core [2].
>
> These numbers are observations from one TLS / QUIC stack doing
> micro-benchmarks, though I tend to think that sharing performance numbers
> lead to people having better expectations regarding the CPU cost of QUIC.
> Some of us were also concerned about the cost of header protection (a.k.a=
.
> packet number encryption) [4]. Hence posting here.
>
> [1] https://github.com/h2o/quicly
> [2] https://github.com/h2o/quicly/pull/359
> [3] https://github.com/h2o/picotls/pull/310
> [4]
> https://mailarchive.ietf.org/arch/msg/quic/Wy8JXwKyeYgWgRWizWDbyOLg0xc/
>
> --
> Kazuho Oku
>

--00000000000056d08905a85f35d7
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<html><head><style>body{font-family:Helvetica,Arial;font-size:13px}</style>=
</head><body><div style=3D"font-family:Helvetica,Arial;font-size:13px">I ag=
ree, thanks for sharing. I am not too surprised by the results though.</div=
><div style=3D"font-family:Helvetica,Arial;font-size:13px"><br></div><div s=
tyle=3D"font-family:Helvetica,Arial;font-size:13px">I am a bit surprised th=
at it is possible to use bulk encrypt headers as Nick suggests as I would h=
ave thought the key / nonce was associated with the packet, but I haven=E2=
=80=99t looked closely recently.</div> <br> <div class=3D"gmail_signature">=
<div style=3D"font-family:helvetica,arial;font-size:13px">Kind Regards,</di=
v><div style=3D"font-family:helvetica,arial;font-size:13px">Mikkel Fahn=C3=
=B8e J=C3=B8rgensen<br><br></div></div> <br><p class=3D"airmail_on">On 18 J=
une 2020 at 19.21.44, Matt Joras (<a href=3D"mailto:matt.joras@gmail.com">m=
att.joras@gmail.com</a>) wrote:</p> <blockquote type=3D"cite" class=3D"clea=
n_bq"><span><div><div></div><div><div dir=3D"ltr"><div><div><div>Wow, reall=
y interesting results, thanks for sharing!<br><br></div>I was curious, sinc=
e it&#39;s not mentioned on the PRs that I saw, were these tests done with =
any adjustments to the ACK frequency on the client? Or is this with the def=
ault ACK policy of ACKing every other?<br><br></div></div>Matt Joras<br></d=
iv><br><div class=3D"gmail_quote"><div dir=3D"ltr" class=3D"gmail_attr">On =
Thu, Jun 18, 2020 at 1:24 AM Kazuho Oku &lt;<a href=3D"mailto:kazuhooku@gma=
il.com">kazuhooku@gmail.com</a>&gt; wrote:<br></div><blockquote class=3D"gm=
ail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,=
204,204);padding-left:1ex"><div dir=3D"ltr"><div dir=3D"ltr"><div dir=3D"lt=
r"><div dir=3D"ltr"><div dir=3D"ltr">Hello folks,<div><br></div><div>TL;DR:=
 When existing crypto libraries are used, the AEAD cost of QUIC can become =
somewhat greater than that of TLS over TCP, due to those crypto libraries n=
ot optimized for handling short AEAD blocks. Crypto libraries optimized for=
=C2=A0short AEAD blocks can tighten the gap. Also, crypto libraries optimiz=
ed for QUIC can combine AEAD operation and header protection vector generat=
ion, effectively eliminating the cost of header protection.</div><div><br><=
/div><div>A bit longer story:</div><div><br></div><div>Recently, when testi=
ng quicly=C2=A0[1] using a NIC with *hardware support* for UDP GSO offload,=
 we noticed that it uses +30% CPU cycles compared to TLS over TCP. This was=
 due to our underlying crypto not being optimized for short AEAD blocks, an=
d due to the call overhead of doing header protection one packet=C2=A0at a =
time. [2]</div><div><br></div><div>We went on to implement our own AES-GCM =
library optimized for short AEAD blocks (i.e. packet-sized), that also gene=
rates the header protection vector while doing AEAD computation. By switchi=
ng to the new crypto library, the performance drawback of the crypto itself=
 compared to large AEAD blocks when down from 30% to 10% [3].=C2=A0</div><d=
iv><br></div><div>The overhead of header protection (relative to all the cr=
ypto cost) on the send side is now between 0.3% to 1% when using packets si=
zed around 1400 bytes to 1500 bytes. The ratio against the total transport =
cost is going to be less than that, and in practice, I think that they are =
negligible. The story might be a bit different on the receive-side due to t=
he dependency between header protection and AEAD, though I am sure there wo=
uld also be ways to optimize.</div><div><br></div><div>In terms of CPU cycl=
es spent by the QUIC stack, the overhead compared to TLS-over-TCP came down=
 to 5%. quicly can now sustain the send rate of 5Gbps while spending about =
30% of CPU cycles of one modern Intel CPU core [2].</div><div><br></div><di=
v>These numbers are observations from one TLS / QUIC stack doing micro-benc=
hmarks, though I tend to think that sharing performance numbers lead to peo=
ple having better expectations regarding the CPU cost of QUIC. Some of us w=
ere also concerned about the cost of header protection (a.k.a. packet numbe=
r encryption) [4]. Hence posting here.<br clear=3D"all"><div><br></div><div=
>[1] <a href=3D"https://github.com/h2o/quicly" target=3D"_blank">https://gi=
thub.com/h2o/quicly</a></div><div><span style=3D"color:rgb(0,0,0)">[2]=C2=
=A0<a href=3D"https://github.com/h2o/quicly/pull/359" target=3D"_blank">htt=
ps://github.com/h2o/quicly/pull/359</a></span></div><div><span style=3D"col=
or:rgb(0,0,0)"></span>[3]=C2=A0<a href=3D"https://github.com/h2o/picotls/pu=
ll/310" target=3D"_blank">https://github.com/h2o/picotls/pull/310</a></div>=
<div>[4]=C2=A0<a href=3D"https://mailarchive.ietf.org/arch/msg/quic/Wy8JXwK=
yeYgWgRWizWDbyOLg0xc/" target=3D"_blank">https://mailarchive.ietf.org/arch/=
msg/quic/Wy8JXwKyeYgWgRWizWDbyOLg0xc/</a></div><div><br></div>-- <br><div d=
ir=3D"ltr">Kazuho Oku</div></div></div></div></div></div></div>
</blockquote></div>
</div></div></span></blockquote></body></html>

--00000000000056d08905a85f35d7--


From nobody Thu Jun 18 14:40:21 2020
Return-Path: <kazuhooku@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 3ECA33A0FD5 for <quic@ietfa.amsl.com>; Thu, 18 Jun 2020 14:40:20 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id fB54K1SmN8Dx for <quic@ietfa.amsl.com>; Thu, 18 Jun 2020 14:40:17 -0700 (PDT)
Received: from mail-ej1-x62c.google.com (mail-ej1-x62c.google.com [IPv6:2a00:1450:4864:20::62c]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 7FF283A0FCC for <quic@ietf.org>; Thu, 18 Jun 2020 14:40:17 -0700 (PDT)
Received: by mail-ej1-x62c.google.com with SMTP id gl26so7976701ejb.11 for <quic@ietf.org>; Thu, 18 Jun 2020 14:40:17 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=tYpt/bLFhuMwzYQRbKsPu6AeSeiCN5C1j4JBhTXdWjM=; b=d6JcVEIw8reKjAEqPxHnCAYghfi4cSeNMeAEGresdB+pj5X88Bb9JFzLfR29rUpaaW ZWdJdcuqXIsFC6Gsq7hefCf99N/O83A+RACoj7eqT0qxWxiTVg+GwhFmVj/jGxk6HeF1 C/M9EZz4fPBweSdcK6+tQJ+WeWIMCxt/od/yT7M5Lgp75wGDia+Mtq1oF4Rsnxixhr7d zcgBwwgK+wv/4+E3EMy46Hk37BCj3Y7O86Hzv66ERvIMdLrOR0X9ArMTEToNx08RE4xY mkTa7FDTJlz4bpZFHRMmxG3+sdmz8HgDog+Mw/cDdxdiyQ5wiURh/lVE44/ecxEdlA/M VQHQ==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=tYpt/bLFhuMwzYQRbKsPu6AeSeiCN5C1j4JBhTXdWjM=; b=aDtHRw1EjK2zgoQ6OVvoZ1+aoVpqzpHAVYNHRxtc8iQm8yJ1yTlfZhexLCDQk6p1Ri UkOTERUXp9z6FJrLUPrlX5D63isRmFQqkUJGuWm/K2cyQCxz9nXRHodsoe5qjCOwMq0m NN2tBAfnvmuksYp0KpUq1p43OHF0skQ2AKtQiUp3p0zRi2Dw93VtHLXKlUbfryoD+K8P ZeWTO2B5RRgKF+yggaK9CbNnNAOtVN6jNNxMHOhVg4GqKG2GFzkDOE0sa/xwEendgmAX XJj+EvOiICxvwQ+44B8nTcajQKbrCIoQBTzdFb2327o/1KBU+gsFV3Hb0/BW2KGAEDvc jdqA==
X-Gm-Message-State: AOAM533/fny8qNEtUsG9xL1hcw1AIG72fWhcEzmFw+2CWrSDCKhb7QLB fK9ZFiHH3Q99dQI+kqK5W8sMLD3qTXdNpGf2bOI=
X-Google-Smtp-Source: ABdhPJwQLCb109T/1u44KKBkEIl9MPCNDJcwTbBdHnk+cv9SIVJlXeJSTQPBWarnCsCG/VYBjSM9mOri5eqF4N8yxe0=
X-Received: by 2002:a17:906:a01:: with SMTP id w1mr695125ejf.197.1592516403654;  Thu, 18 Jun 2020 14:40:03 -0700 (PDT)
MIME-Version: 1.0
References: <CANatvzz8F1H=DXMkBEhmKHnYM-HVG48TS9KwY=OP881Txkcodw@mail.gmail.com> <CADdTf+i+LZ98GgNhFNVcuoVczC=jCQE-TqWbCqhrpR7=Z2knWg@mail.gmail.com> <CAN1APdft3UU1dfKY_UxRaLy2xeCYSXQT3=k53=96OO_Gu1X_cw@mail.gmail.com>
In-Reply-To: <CAN1APdft3UU1dfKY_UxRaLy2xeCYSXQT3=k53=96OO_Gu1X_cw@mail.gmail.com>
From: Kazuho Oku <kazuhooku@gmail.com>
Date: Fri, 19 Jun 2020 06:39:52 +0900
Message-ID: <CANatvzw2pj-CaXbAJ_-kUrvmoC3_oyvnSo7Yn+mX-kBuqVkr3w@mail.gmail.com>
Subject: Re: AEAD and header encryption overhead in QUIC
To: =?UTF-8?Q?Mikkel_Fahn=C3=B8e_J=C3=B8rgensen?= <mikkelfj@gmail.com>,  Nick Banks <nibanks@microsoft.com>
Cc: Matt Joras <matt.joras@gmail.com>, IETF QUIC WG <quic@ietf.org>
Content-Type: multipart/alternative; boundary="00000000000089f54805a8629ec3"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/0u-hWb32SlqzEgAvu7-2Tuatl_w>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 18 Jun 2020 21:40:20 -0000

--00000000000089f54805a8629ec3
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Nick, Matt, Mikkel, thank you for your comments.

2020=E5=B9=B46=E6=9C=8818=E6=97=A5(=E6=9C=A8) 23:52 Nick Banks <nibanks@mic=
rosoft.com>:

> We=E2=80=99ve found that an easy way to minimize the CPU cost of the head=
er
> protection (on send and receive) is batching. If you copy 8 packet header=
s
> into a single contiguous block of memory, you can do a single crypto
> operation on it. The cost of doing one batch of 8 is essentially the same
> as doing just a single header. This effectively cuts the cost of header
> protection to 1/8th the original cost. Obviously this only works if you
> have enough packets to batch process, but for high throughput tests, you
> usually do.
>
>
>
> Feel free to take a look at the MsQuic code (receive path
> <https://github.com/microsoft/msquic/blob/master/src/core/connection.c#L4=
784>
> , send path
> <https://github.com/microsoft/msquic/blob/master/src/core/packet_builder.=
c#L728>).
> The nice thing about this design (for those of us who don=E2=80=99t have =
as much
> special crypto expertise) is that it doesn=E2=80=99t require any special =
crypto
> changes. I=E2=80=99d be interested to see if our two approaches could be =
combined
> for even better performance!
>

Thank you for sharing your experience. This is indeed a sensible approach
for minimizing the cost of header protection. On the receive side,
processing multiple packets in batch is the only way of minimizing the cost
of header protection, because AEAD unprotection cannot start until the
result of header unprotection is being obtained. Considering that modern
x86-64 CPUs can run 4 to 8 AES block operations in parallel, doing header
protection for one packet at a time is a waste of CPU resources.

To paraphrase, the two approaches are orthogonal on the receive-side, and
combining them would give better performance. My napkin tells me that I can
expect at most 3% reduction in CPU cycles spent in crypto.

On the send side however, I do not expect to see performance increase by
combining the two approaches. This "game" is about keeping the CPU
pipelines that do AES-NI running at their full speed. The cost does not
change as long as the AES operations of header protection are run in
parallel with other AES operations, regardless of that being AEAD or header
protection o other packets.

If there's chance of further improving performance, I tend to think that
that would come from overlapping the AEAD operation of multiple packets.
This "game" is about keeping the AES-NI pipeline busy. With Fusion, we've
reached 90% of the theoretical maximum speed (9th-gen Intel Core CPUs can
do 64 bytes of AES encryption in 40 clocks). The question to us is if we
want to consider changing our API for single-digit performance gain.


2020=E5=B9=B46=E6=9C=8819=E6=97=A5(=E9=87=91) 2:21 Matt Joras <matt.joras@g=
mail.com>:

> I was curious, since it's not mentioned on the PRs that I saw, were these
> tests done with any adjustments to the ACK frequency on the client? Or is
> this with the default ACK policy of ACKing every other?
>

This is an important point, thank you for asking. We did reduce ACK
frequency to 1/8 CWND, as we did in our previous report [1]. In fact, you
cannot turn that off with quicly.


2020=E5=B9=B46=E6=9C=8819=E6=97=A5(=E9=87=91) 2:35 Mikkel Fahn=C3=B8e J=C3=
=B8rgensen <mikkelfj@gmail.com>:

> I am a bit surprised that it is possible to use bulk encrypt headers as
> Nick suggests as I would have thought the key / nonce was associated with
> the packet, but I haven=E2=80=99t looked closely recently.
>

That's a keen observation. Nonce used for header protection does depend on
the output of AEAD. Though, because the nonce is taken from the very first
few bytes of AEAD output, it is possible to start calculating the header
protection vector before finishing AEAD operation, unless the packet is
tiny. The position of the nonce was deliberately chosen to provide room for
this type of optimization. I recall at least one person (Martin Thomson)
arguing for having this possibility.

[1]
https://www.fastly.com/blog/measuring-quic-vs-tcp-computational-efficiency

--=20
Kazuho Oku

--00000000000089f54805a8629ec3
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div dir=3D"ltr"><div dir=3D"ltr"><div dir=3D"ltr"><div di=
r=3D"ltr"><div dir=3D"ltr"><div dir=3D"ltr"><div dir=3D"ltr">Nick, Matt, Mi=
kkel, thank you for your comments.<div><br></div><div><span style=3D"color:=
rgb(0,0,0)">2020=E5=B9=B46=E6=9C=8818=E6=97=A5(=E6=9C=A8) 23:52 Nick Banks =
&lt;<a href=3D"mailto:nibanks@microsoft.com">nibanks@microsoft.com</a>&gt;:=
</span><br></div><div><blockquote class=3D"gmail_quote" style=3D"color:rgb(=
0,0,0);margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-style:sol=
id;border-left-color:rgb(204,204,204);padding-left:1ex"><div lang=3D"EN-US"=
><div class=3D"gmail-m_8001779623976103357WordSection1"><p class=3D"MsoNorm=
al">We=E2=80=99ve found that an easy way to minimize the CPU cost of the he=
ader protection (on send and receive) is batching. If you copy 8 packet hea=
ders into a single contiguous block of memory, you can do a single crypto o=
peration on it. The cost of doing one batch of 8 is essentially the same as=
 doing just a single header. This effectively cuts the cost of header prote=
ction to 1/8<sup>th</sup>=C2=A0the original cost. Obviously this only works=
 if you have enough packets to batch process, but for high throughput tests=
, you usually do.<br></p><p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p><p =
class=3D"MsoNormal">Feel free to take a look at the MsQuic code (<a href=3D=
"https://github.com/microsoft/msquic/blob/master/src/core/connection.c#L478=
4" target=3D"_blank">receive path</a>,=C2=A0<a href=3D"https://github.com/m=
icrosoft/msquic/blob/master/src/core/packet_builder.c#L728" target=3D"_blan=
k">send path</a>). The nice thing about this design (for those of us who do=
n=E2=80=99t have as much special crypto expertise) is that it doesn=E2=80=
=99t require any special crypto changes. I=E2=80=99d be interested to see i=
f our two approaches could be combined for even better performance!</p></di=
v></div></blockquote><div><br></div><div>Thank you for sharing your experie=
nce. This is indeed a sensible approach for minimizing the cost of header p=
rotection. On the receive side, processing multiple packets in batch is the=
 only way of minimizing the cost of header protection, because AEAD unprote=
ction cannot start until the result of header unprotection is being obtaine=
d. Considering that modern x86-64 CPUs can run 4 to 8 AES block operations =
in parallel, doing header protection for one packet at a time is a waste of=
 CPU resources.</div><div><br></div><div>To paraphrase, the two approaches =
are orthogonal on the receive-side, and combining them would give better pe=
rformance. My napkin tells me that I can expect at most 3% reduction in CPU=
 cycles spent in crypto.</div><div><br></div><div>On the send side however,=
 I do not expect to see performance increase by combining the two approache=
s. This &quot;game&quot; is about keeping the CPU pipelines that do AES-NI =
running at their full speed. The cost does not change as long as the AES op=
erations of header protection are run in parallel with other AES operations=
, regardless of that being AEAD or header protection o other packets.</div>=
<div><br></div><div>If there&#39;s chance of further improving performance,=
 I tend to think that that would come from overlapping the AEAD operation o=
f multiple packets. This &quot;game&quot; is about keeping the AES-NI pipel=
ine busy. With Fusion, we&#39;ve reached 90% of the theoretical maximum spe=
ed (9th-gen Intel Core CPUs can do 64 bytes of AES encryption in 40 clocks)=
. The question to us is if we want to consider changing our API for single-=
digit performance gain.</div><div>=C2=A0</div></div><div><br></div><div><di=
v dir=3D"ltr" class=3D"gmail_attr" style=3D"color:rgb(0,0,0)">2020=E5=B9=B4=
6=E6=9C=8819=E6=97=A5(=E9=87=91) 2:21 Matt Joras &lt;<a href=3D"mailto:matt=
.joras@gmail.com">matt.joras@gmail.com</a>&gt;:</div><blockquote class=3D"g=
mail_quote" style=3D"color:rgb(0,0,0);margin:0px 0px 0px 0.8ex;border-left-=
width:1px;border-left-style:solid;border-left-color:rgb(204,204,204);paddin=
g-left:1ex"><div dir=3D"ltr">I was curious, since it&#39;s not mentioned on=
 the PRs that I saw, were these tests done with any adjustments to the ACK =
frequency on the client? Or is this with the default ACK policy of ACKing e=
very other?</div></blockquote><div><br></div><div>This is an important poin=
t, thank you for asking. We did reduce ACK frequency to 1/8 CWND, as we did=
 in our previous report [1]. In fact, you cannot turn that off with quicly.=
</div></div><div><br></div><div><br></div></div></div></div></div></div></d=
iv><div class=3D"gmail_quote"><div dir=3D"ltr" class=3D"gmail_attr">2020=E5=
=B9=B46=E6=9C=8819=E6=97=A5(=E9=87=91) 2:35 Mikkel Fahn=C3=B8e J=C3=B8rgens=
en &lt;<a href=3D"mailto:mikkelfj@gmail.com">mikkelfj@gmail.com</a>&gt;:</d=
iv><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;bord=
er-left-width:1px;border-left-style:solid;border-left-color:rgb(204,204,204=
);padding-left:1ex"><div><div style=3D"font-family:Helvetica,Arial;font-siz=
e:13px">I am a bit surprised that it is possible to use bulk encrypt header=
s as Nick suggests as I would have thought the key / nonce was associated w=
ith the packet, but I haven=E2=80=99t looked closely recently.</div> </div>=
</blockquote></div><div><br></div><div>That&#39;s a keen observation. Nonce=
 used for header protection does depend on the output of AEAD. Though, beca=
use the nonce is taken from the very first few bytes of AEAD output, it is =
possible to start calculating the header protection vector before finishing=
 AEAD operation, unless the packet is tiny. The position of the nonce was d=
eliberately chosen to provide room for this type of optimization. I recall =
at least one person (Martin Thomson) arguing for having this possibility.</=
div><div><br></div><div>[1]=C2=A0<a href=3D"https://www.fastly.com/blog/mea=
suring-quic-vs-tcp-computational-efficiency">https://www.fastly.com/blog/me=
asuring-quic-vs-tcp-computational-efficiency</a></div><div><br></div>-- <br=
><div dir=3D"ltr" class=3D"gmail_signature">Kazuho Oku</div></div></div>

--00000000000089f54805a8629ec3--


From nobody Thu Jun 18 16:20:01 2020
Return-Path: <mnot@mnot.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id A4E923A1125 for <quic@ietfa.amsl.com>; Thu, 18 Jun 2020 16:19:59 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.098
X-Spam-Level: 
X-Spam-Status: No, score=-2.098 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_MSPIKE_H4=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=mnot.net header.b=XmeHg3nQ; dkim=pass (2048-bit key) header.d=messagingengine.com header.b=I8hYC837
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id nxR-mOeEegwV for <quic@ietfa.amsl.com>; Thu, 18 Jun 2020 16:19:57 -0700 (PDT)
Received: from out3-smtp.messagingengine.com (out3-smtp.messagingengine.com [66.111.4.27]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 9D6FA3A111C for <quic@ietf.org>; Thu, 18 Jun 2020 16:19:57 -0700 (PDT)
Received: from compute4.internal (compute4.nyi.internal [10.202.2.44]) by mailout.nyi.internal (Postfix) with ESMTP id 045DC5C0163; Thu, 18 Jun 2020 19:19:56 -0400 (EDT)
Received: from mailfrontend2 ([10.202.2.163]) by compute4.internal (MEProxy); Thu, 18 Jun 2020 19:19:56 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=mnot.net; h= content-type:mime-version:subject:from:in-reply-to:date:cc :content-transfer-encoding:message-id:references:to; s=fm3; bh=m l49vMo98qUlevgPEIOsfPoBc88j18T39UpnYFstZrk=; b=XmeHg3nQaOZCyQTQp INfGlWvOPv8T/xIkFfgDN9rI6NV5YDDyqnHb3aRE4+3PbrJQXxuS28+pCzUnBwm0 UsPc+z0mL5DMUCqk9SRaOC+Rtb7y6trW4gWxOW6dirfv0o7sBI9OAA1ATadCj1mw 82imBMshrGR9medMRHA9JBExOf0P74KBDWvgVQH4yyTVWZ2g0/EJsDusPkx5lqkY emLhegm4VvFYBhMglTn8W6OONFzyPb6dsrHM9ABcF+zg2F0u3IUG//hnWTkw9nh6 ytiLnEtIAT/N01IPolbfVla/NHUYHQH67wRiMvUwGv/B4HjFlKC8LINe/4btJP1b uzd3A==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d= messagingengine.com; h=cc:content-transfer-encoding:content-type :date:from:in-reply-to:message-id:mime-version:references :subject:to:x-me-proxy:x-me-proxy:x-me-sender:x-me-sender :x-sasl-enc; s=fm3; bh=ml49vMo98qUlevgPEIOsfPoBc88j18T39UpnYFstZ rk=; b=I8hYC8378pLCcXqzD1Os/pMX25NgY5gOxlNL0vFKx7xBQGMm2d5PMGkju WUBJ3yvKYGzyPKVyg3Yu7QZ1MxXyOa0sEvFfi5fDwP9BeHL73fVkiScgDs6tESIm fBQHO5VvEqj4xOVS66+FODUb5fyp+6reV6WQlcMDoVXnBOCHImEt1aOXwey4Low0 j67kk5wR22Qv+cAxm/PuSHv4dRPgha01NBVneJbNYd2DDHgCKGNRCs7JAO/3UBFs 2bxraCULTt+in1yqJWolFzq4LCOZW199lpzd1auPi6ch08uNBedl8OTGME5MlJXj yr3H0SR1DIm8rbcM0TkxL3RpqWcWA==
X-ME-Sender: <xms:mfbrXuF66e7LUP2eF3wgp9yF64c0HqnIMiBThoLVOEaBAKUMM9QoYQ>
X-ME-Proxy-Cause: gggruggvucftvghtrhhoucdtuddrgeduhedrudejhedgvdduucetufdoteggodetrfdotf fvucfrrhhofhhilhgvmecuhfgrshhtofgrihhlpdfqfgfvpdfurfetoffkrfgpnffqhgen uceurghilhhouhhtmecufedttdenucesvcftvggtihhpihgvnhhtshculddquddttddmne cujfgurheptggguffhjgffgffkfhfvofesthhqmhdthhdtvdenucfhrhhomhepofgrrhhk ucfpohhtthhinhhghhgrmhcuoehmnhhothesmhhnohhtrdhnvghtqeenucggtffrrghtth gvrhhnpeevffffhfduteevvefhueffieegtdeutdehffeltefffedttdeggeejheeiueet teenucffohhmrghinhepmhhnohhtrdhnvghtnecukfhppeduudelrddujedrudehkedrvd ehudenucevlhhushhtvghrufhiiigvpedtnecurfgrrhgrmhepmhgrihhlfhhrohhmpehm nhhothesmhhnohhtrdhnvght
X-ME-Proxy: <xmx:mfbrXvWTMSv8l6EtdbrO21PyWwK-_tryRRVOk-_I4flt7ii4OdyEZA> <xmx:mfbrXoJ1RsKbPU355cvsTSRiyo_4c-pqHtgk0qm1jRRquF-pnizCNw> <xmx:mfbrXoFATlrnIkCnKCGcHUFUGaY6c3T8dfVLcBGCiZmRGNNJkPvaOQ> <xmx:m_brXvd6IstVFozbTrVIK2NT6TnjJPmyBCig50BNIUAEQhQWtH1H9g>
Received: from macbook-air.mnot.net (119-17-158-251.77119e.mel.static.aussiebb.net [119.17.158.251]) by mail.messagingengine.com (Postfix) with ESMTPA id B55563061CB6; Thu, 18 Jun 2020 19:19:52 -0400 (EDT)
Content-Type: text/plain; charset=us-ascii
Mime-Version: 1.0 (Mac OS X Mail 13.4 \(3608.80.23.2.2\))
Subject: Re: draft-ietf-quic-http-29, "1.1. Prior versions of HTTP"
From: Mark Nottingham <mnot@mnot.net>
In-Reply-To: <8c747b59-1672-4ccb-dfce-0914e74c504f@gmx.de>
Date: Fri, 19 Jun 2020 09:19:48 +1000
Cc: quic@ietf.org
Content-Transfer-Encoding: quoted-printable
Message-Id: <9B975808-6BF3-491C-8E32-C475824B0E9A@mnot.net>
References: <8c747b59-1672-4ccb-dfce-0914e74c504f@gmx.de>
To: "Julian F. Reschke" <julian.reschke@gmx.de>
X-Mailer: Apple Mail (2.3608.80.23.2.2)
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/TmD4CppstBUqF4KYuQCLMSyIQvA>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 18 Jun 2020 23:20:00 -0000

My .02 -

I think we need to balance between being correct and over-complicating =
the text here. In *theory*, Julian is correct; in wide practice, =
pipelining isn't used because of the limitations it has, as well as =
implementation / interop issues.

How about something like:

"Because of limitations in how HTTP exchanges are mapped onto TCP =
connections in HTTP/1.1, multiple parallel connections are often =
used..."

?

Cheers,


> On 12 Jun 2020, at 9:48 pm, Julian Reschke <julian.reschke@gmx.de> =
wrote:
>=20
>> HTTP/1.1 [HTTP11] is a TCP mapping which uses whitespace-delimited =
text fields to convey HTTP messages. While these exchanges are =
human-readable, using whitespace for message formatting leads to parsing =
complexity and motivates tolerance of variant behavior. Because each =
connection can transfer only a single HTTP request or response at a time =
in each direction, multiple parallel TCP connections are often used, =
reducing the ability of the congestion controller to effectively manage =
traffic between endpoints.
>=20
> Isn't that a bit misleading, as at least in theory HTTP/1.1 supports
> pipelining?
>=20
> BEst regards, Julian
>=20

--
Mark Nottingham   https://www.mnot.net/


From nobody Thu Jun 18 16:24:35 2020
Return-Path: <matt.joras@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 104CF3A0028 for <quic@ietfa.amsl.com>; Thu, 18 Jun 2020 16:24:34 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.096
X-Spam-Level: 
X-Spam-Status: No, score=-2.096 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_BLOCKED=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id RcYZ5Vc4uzZE for <quic@ietfa.amsl.com>; Thu, 18 Jun 2020 16:24:32 -0700 (PDT)
Received: from mail-vk1-xa2b.google.com (mail-vk1-xa2b.google.com [IPv6:2607:f8b0:4864:20::a2b]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id EC0833A1219 for <quic@ietf.org>; Thu, 18 Jun 2020 16:24:31 -0700 (PDT)
Received: by mail-vk1-xa2b.google.com with SMTP id s192so1874268vkh.3 for <quic@ietf.org>; Thu, 18 Jun 2020 16:24:31 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=B+RwLoo9cOjKN+Las/AgPeza930ZmE1gGI5M/0clJGY=; b=d9fXmBHrDmMY/iSIxRzJg76Wa1px9UUdv0X2nvrgytCsemKo7qmxi8hx6e22WfyuKW BkpvjGEhx5UhsAtf91jgaKqAF0z7DL5P8asmjJh5vCU5nAh3p++3pNLnbgMPOQN94SaM FwqSIRAqDJGG/GkmeybLd1LgUcRKAyp5++FzvHV39kQzWPNXdVnI0SkjvpA2/DN8c8U6 28WH+D/mj/zQHIHVh07BUiDzNIbwX+0bZHuoIAP77rDGQmrLYQtAUbM9Zl+g8d66kvF3 q2R+9LUtg8kzoglFJ4glgj9RdGj32jsKquKBMlbVeJGW+Iat38mHCtjj3duHe4p/toqC bXbA==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=B+RwLoo9cOjKN+Las/AgPeza930ZmE1gGI5M/0clJGY=; b=YY3ER3QoeWYEQB2PyOVT7rAuti0gCLjR7NS2hheAWjZ468LuT6aNZX3rFJ1oHW8ci+ CL+fulDIT7sGTR4yxE4nrWsstelZ3PK9oSbTY7+CwPrqz/kzEvGdIDrOnW5QXSuuAlkL mcjil1pk/U80s2Q9f71/L9VzA6K6rGDKKSSr8sqyol4VvpCXHgjHvusMnXekLb+oYUwk dkZu1b1BQsAPrIbVawShnCGwQOr1BoyLc/uX/w3T/8qtO4XAU2Dm6d4Vm65bfgchfUrh r9zGTdhsUYOcz3h7nO2FPuESVCI47Q1eb7G3xHKL1UUaOAzBLJTeTZpArLx1K4cKyj6/ +Uvw==
X-Gm-Message-State: AOAM531Ef7jDuTQxklAJdSzDW6BCcmYcTattRvXnG5GtB9nNXYZIhYsO VQkaUgrQiOJwB4XFXL+8ApC/bQiyqvPJlMqJuwg=
X-Google-Smtp-Source: ABdhPJyb0WGxR/X4CACR4XTA6v0FtXsTyFX+D55mdO/e3LfTJki7I6eurpwZcJ9+u0HCZW4IGwLopFovYgbMkMhz/Uw=
X-Received: by 2002:a05:6122:10f8:: with SMTP id m24mr5404576vko.41.1592522670639;  Thu, 18 Jun 2020 16:24:30 -0700 (PDT)
MIME-Version: 1.0
References: <CANatvzz8F1H=DXMkBEhmKHnYM-HVG48TS9KwY=OP881Txkcodw@mail.gmail.com> <CADdTf+i+LZ98GgNhFNVcuoVczC=jCQE-TqWbCqhrpR7=Z2knWg@mail.gmail.com> <CAN1APdft3UU1dfKY_UxRaLy2xeCYSXQT3=k53=96OO_Gu1X_cw@mail.gmail.com> <CANatvzw2pj-CaXbAJ_-kUrvmoC3_oyvnSo7Yn+mX-kBuqVkr3w@mail.gmail.com>
In-Reply-To: <CANatvzw2pj-CaXbAJ_-kUrvmoC3_oyvnSo7Yn+mX-kBuqVkr3w@mail.gmail.com>
From: Matt Joras <matt.joras@gmail.com>
Date: Thu, 18 Jun 2020 16:24:19 -0700
Message-ID: <CADdTf+jtAd5rh+fW2RMLjTgXzNZ30xxSQMXKRtRb4z7bZcRsbg@mail.gmail.com>
Subject: Re: AEAD and header encryption overhead in QUIC
To: Kazuho Oku <kazuhooku@gmail.com>
Cc: =?UTF-8?Q?Mikkel_Fahn=C3=B8e_J=C3=B8rgensen?= <mikkelfj@gmail.com>,  Nick Banks <nibanks@microsoft.com>, IETF QUIC WG <quic@ietf.org>
Content-Type: multipart/alternative; boundary="000000000000148c6c05a8641438"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/5c9Y3-ntudAwRLk2H96Vag6efBw>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 18 Jun 2020 23:24:34 -0000

--000000000000148c6c05a8641438
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Thanks for the explanations, Kazuho! Would it be possible to share the
commands you used to produce the given results? It would be interesting so
others can do direct comparisons with their own performance benchmarking.
It is always hard to do comparisons across implementations with varying
hardware and OS versions.

Matt Joras

On Thu, Jun 18, 2020 at 2:40 PM Kazuho Oku <kazuhooku@gmail.com> wrote:

> Nick, Matt, Mikkel, thank you for your comments.
>
> 2020=E5=B9=B46=E6=9C=8818=E6=97=A5(=E6=9C=A8) 23:52 Nick Banks <nibanks@m=
icrosoft.com>:
>
>> We=E2=80=99ve found that an easy way to minimize the CPU cost of the hea=
der
>> protection (on send and receive) is batching. If you copy 8 packet heade=
rs
>> into a single contiguous block of memory, you can do a single crypto
>> operation on it. The cost of doing one batch of 8 is essentially the sam=
e
>> as doing just a single header. This effectively cuts the cost of header
>> protection to 1/8th the original cost. Obviously this only works if you
>> have enough packets to batch process, but for high throughput tests, you
>> usually do.
>>
>>
>>
>> Feel free to take a look at the MsQuic code (receive path
>> <https://github.com/microsoft/msquic/blob/master/src/core/connection.c#L=
4784>
>> , send path
>> <https://github.com/microsoft/msquic/blob/master/src/core/packet_builder=
.c#L728>).
>> The nice thing about this design (for those of us who don=E2=80=99t have=
 as much
>> special crypto expertise) is that it doesn=E2=80=99t require any special=
 crypto
>> changes. I=E2=80=99d be interested to see if our two approaches could be=
 combined
>> for even better performance!
>>
>
> Thank you for sharing your experience. This is indeed a sensible approach
> for minimizing the cost of header protection. On the receive side,
> processing multiple packets in batch is the only way of minimizing the co=
st
> of header protection, because AEAD unprotection cannot start until the
> result of header unprotection is being obtained. Considering that modern
> x86-64 CPUs can run 4 to 8 AES block operations in parallel, doing header
> protection for one packet at a time is a waste of CPU resources.
>
> To paraphrase, the two approaches are orthogonal on the receive-side, and
> combining them would give better performance. My napkin tells me that I c=
an
> expect at most 3% reduction in CPU cycles spent in crypto.
>
> On the send side however, I do not expect to see performance increase by
> combining the two approaches. This "game" is about keeping the CPU
> pipelines that do AES-NI running at their full speed. The cost does not
> change as long as the AES operations of header protection are run in
> parallel with other AES operations, regardless of that being AEAD or head=
er
> protection o other packets.
>
> If there's chance of further improving performance, I tend to think that
> that would come from overlapping the AEAD operation of multiple packets.
> This "game" is about keeping the AES-NI pipeline busy. With Fusion, we've
> reached 90% of the theoretical maximum speed (9th-gen Intel Core CPUs can
> do 64 bytes of AES encryption in 40 clocks). The question to us is if we
> want to consider changing our API for single-digit performance gain.
>
>
> 2020=E5=B9=B46=E6=9C=8819=E6=97=A5(=E9=87=91) 2:21 Matt Joras <matt.joras=
@gmail.com>:
>
>> I was curious, since it's not mentioned on the PRs that I saw, were thes=
e
>> tests done with any adjustments to the ACK frequency on the client? Or i=
s
>> this with the default ACK policy of ACKing every other?
>>
>
> This is an important point, thank you for asking. We did reduce ACK
> frequency to 1/8 CWND, as we did in our previous report [1]. In fact, you
> cannot turn that off with quicly.
>
>
> 2020=E5=B9=B46=E6=9C=8819=E6=97=A5(=E9=87=91) 2:35 Mikkel Fahn=C3=B8e J=
=C3=B8rgensen <mikkelfj@gmail.com>:
>
>> I am a bit surprised that it is possible to use bulk encrypt headers as
>> Nick suggests as I would have thought the key / nonce was associated wit=
h
>> the packet, but I haven=E2=80=99t looked closely recently.
>>
>
> That's a keen observation. Nonce used for header protection does depend o=
n
> the output of AEAD. Though, because the nonce is taken from the very firs=
t
> few bytes of AEAD output, it is possible to start calculating the header
> protection vector before finishing AEAD operation, unless the packet is
> tiny. The position of the nonce was deliberately chosen to provide room f=
or
> this type of optimization. I recall at least one person (Martin Thomson)
> arguing for having this possibility.
>
> [1]
> https://www.fastly.com/blog/measuring-quic-vs-tcp-computational-efficienc=
y
>
> --
> Kazuho Oku
>

--000000000000148c6c05a8641438
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div>Thanks for the explanations, Kazuho! Would it be poss=
ible to share the commands you used to produce the given results? It would =
be interesting so others can do direct comparisons with their own performan=
ce benchmarking. It is always hard to do comparisons across implementations=
 with varying hardware and OS versions.</div><div><br></div><div>Matt Joras=
<br></div></div><br><div class=3D"gmail_quote"><div dir=3D"ltr" class=3D"gm=
ail_attr">On Thu, Jun 18, 2020 at 2:40 PM Kazuho Oku &lt;<a href=3D"mailto:=
kazuhooku@gmail.com">kazuhooku@gmail.com</a>&gt; wrote:<br></div><blockquot=
e class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px s=
olid rgb(204,204,204);padding-left:1ex"><div dir=3D"ltr"><div dir=3D"ltr"><=
div dir=3D"ltr"><div dir=3D"ltr"><div dir=3D"ltr"><div dir=3D"ltr"><div dir=
=3D"ltr"><div dir=3D"ltr">Nick, Matt, Mikkel, thank you for your comments.<=
div><br></div><div><span style=3D"color:rgb(0,0,0)">2020=E5=B9=B46=E6=9C=88=
18=E6=97=A5(=E6=9C=A8) 23:52 Nick Banks &lt;<a href=3D"mailto:nibanks@micro=
soft.com" target=3D"_blank">nibanks@microsoft.com</a>&gt;:</span><br></div>=
<div><blockquote class=3D"gmail_quote" style=3D"color:rgb(0,0,0);margin:0px=
 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><di=
v lang=3D"EN-US"><div><p class=3D"MsoNormal">We=E2=80=99ve found that an ea=
sy way to minimize the CPU cost of the header protection (on send and recei=
ve) is batching. If you copy 8 packet headers into a single contiguous bloc=
k of memory, you can do a single crypto operation on it. The cost of doing =
one batch of 8 is essentially the same as doing just a single header. This =
effectively cuts the cost of header protection to 1/8<sup>th</sup>=C2=A0the=
 original cost. Obviously this only works if you have enough packets to bat=
ch process, but for high throughput tests, you usually do.<br></p><p class=
=3D"MsoNormal"><u></u>=C2=A0<u></u></p><p class=3D"MsoNormal">Feel free to =
take a look at the MsQuic code (<a href=3D"https://github.com/microsoft/msq=
uic/blob/master/src/core/connection.c#L4784" target=3D"_blank">receive path=
</a>,=C2=A0<a href=3D"https://github.com/microsoft/msquic/blob/master/src/c=
ore/packet_builder.c#L728" target=3D"_blank">send path</a>). The nice thing=
 about this design (for those of us who don=E2=80=99t have as much special =
crypto expertise) is that it doesn=E2=80=99t require any special crypto cha=
nges. I=E2=80=99d be interested to see if our two approaches could be combi=
ned for even better performance!</p></div></div></blockquote><div><br></div=
><div>Thank you for sharing your experience. This is indeed a sensible appr=
oach for minimizing the cost of header protection. On the receive side, pro=
cessing multiple packets in batch is the only way of minimizing the cost of=
 header protection, because AEAD unprotection cannot start until the result=
 of header unprotection is being obtained. Considering that modern x86-64 C=
PUs can run 4 to 8 AES block operations in parallel, doing header protectio=
n for one packet at a time is a waste of CPU resources.</div><div><br></div=
><div>To paraphrase, the two approaches are orthogonal on the receive-side,=
 and combining them would give better performance. My napkin tells me that =
I can expect at most 3% reduction in CPU cycles spent in crypto.</div><div>=
<br></div><div>On the send side however, I do not expect to see performance=
 increase by combining the two approaches. This &quot;game&quot; is about k=
eeping the CPU pipelines that do AES-NI running at their full speed. The co=
st does not change as long as the AES operations of header protection are r=
un in parallel with other AES operations, regardless of that being AEAD or =
header protection o other packets.</div><div><br></div><div>If there&#39;s =
chance of further improving performance, I tend to think that that would co=
me from overlapping the AEAD operation of multiple packets. This &quot;game=
&quot; is about keeping the AES-NI pipeline busy. With Fusion, we&#39;ve re=
ached 90% of the theoretical maximum speed (9th-gen Intel Core CPUs can do =
64 bytes of AES encryption in 40 clocks). The question to us is if we want =
to consider changing our API for single-digit performance gain.</div><div>=
=C2=A0</div></div><div><br></div><div><div dir=3D"ltr" class=3D"gmail_attr"=
 style=3D"color:rgb(0,0,0)">2020=E5=B9=B46=E6=9C=8819=E6=97=A5(=E9=87=91) 2=
:21 Matt Joras &lt;<a href=3D"mailto:matt.joras@gmail.com" target=3D"_blank=
">matt.joras@gmail.com</a>&gt;:</div><blockquote class=3D"gmail_quote" styl=
e=3D"color:rgb(0,0,0);margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(20=
4,204,204);padding-left:1ex"><div dir=3D"ltr">I was curious, since it&#39;s=
 not mentioned on the PRs that I saw, were these tests done with any adjust=
ments to the ACK frequency on the client? Or is this with the default ACK p=
olicy of ACKing every other?</div></blockquote><div><br></div><div>This is =
an important point, thank you for asking. We did reduce ACK frequency to 1/=
8 CWND, as we did in our previous report [1]. In fact, you cannot turn that=
 off with quicly.</div></div><div><br></div><div><br></div></div></div></di=
v></div></div></div><div class=3D"gmail_quote"><div dir=3D"ltr" class=3D"gm=
ail_attr">2020=E5=B9=B46=E6=9C=8819=E6=97=A5(=E9=87=91) 2:35 Mikkel Fahn=C3=
=B8e J=C3=B8rgensen &lt;<a href=3D"mailto:mikkelfj@gmail.com" target=3D"_bl=
ank">mikkelfj@gmail.com</a>&gt;:</div><blockquote class=3D"gmail_quote" sty=
le=3D"margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);paddi=
ng-left:1ex"><div><div style=3D"font-family:Helvetica,Arial;font-size:13px"=
>I am a bit surprised that it is possible to use bulk encrypt headers as Ni=
ck suggests as I would have thought the key / nonce was associated with the=
 packet, but I haven=E2=80=99t looked closely recently.</div> </div></block=
quote></div><div><br></div><div>That&#39;s a keen observation. Nonce used f=
or header protection does depend on the output of AEAD. Though, because the=
 nonce is taken from the very first few bytes of AEAD output, it is possibl=
e to start calculating the header protection vector before finishing AEAD o=
peration, unless the packet is tiny. The position of the nonce was delibera=
tely chosen to provide room for this type of optimization. I recall at leas=
t one person (Martin Thomson) arguing for having this possibility.</div><di=
v><br></div><div>[1]=C2=A0<a href=3D"https://www.fastly.com/blog/measuring-=
quic-vs-tcp-computational-efficiency" target=3D"_blank">https://www.fastly.=
com/blog/measuring-quic-vs-tcp-computational-efficiency</a></div><div><br><=
/div>-- <br><div dir=3D"ltr">Kazuho Oku</div></div></div>
</blockquote></div>

--000000000000148c6c05a8641438--


From nobody Thu Jun 18 20:04:08 2020
Return-Path: <julian.reschke@gmx.de>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 4818C3A10A4 for <quic@ietfa.amsl.com>; Thu, 18 Jun 2020 20:04:07 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, FREEMAIL_FROM=0.001, RCVD_IN_MSPIKE_H2=-0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=gmx.net
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id jwqWfjSN4VLF for <quic@ietfa.amsl.com>; Thu, 18 Jun 2020 20:04:06 -0700 (PDT)
Received: from mout.gmx.net (mout.gmx.net [212.227.15.18]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 7755E3A10A3 for <quic@ietf.org>; Thu, 18 Jun 2020 20:04:05 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=gmx.net; s=badeba3b8450; t=1592535838; bh=OETNzYJs9Ku9LdWJaCzMPQyUChUXwe252kS6gSp5fAA=; h=X-UI-Sender-Class:Subject:To:Cc:References:From:Date:In-Reply-To; b=gswtChiVLy2Lz5vdoppBxVWKEe6oklenjSfCf4EZd6A3LQ9XUErok5sEDc0Uqq+N2 QVwv/eB7MMXQ8oWGugoTn7+hwZU5qiaRNiw+Lv8wfdOST1uOXhPAn57cQuX2rByXcm eLVXZte2i1soCtmZFd0M7ZHwVNx8JaAXhNQMW+7Y=
X-UI-Sender-Class: 01bb95c1-4bf8-414a-932a-4f6e2808ef9c
Received: from [192.168.178.182] ([91.61.49.120]) by mail.gmx.com (mrgmx004 [212.227.17.190]) with ESMTPSA (Nemesis) id 1MWics-1jNxhO28pE-00X4q0; Fri, 19 Jun 2020 05:03:58 +0200
Subject: Re: draft-ietf-quic-http-29, "1.1. Prior versions of HTTP"
To: Mark Nottingham <mnot@mnot.net>
Cc: quic@ietf.org
References: <8c747b59-1672-4ccb-dfce-0914e74c504f@gmx.de> <9B975808-6BF3-491C-8E32-C475824B0E9A@mnot.net>
From: Julian Reschke <julian.reschke@gmx.de>
Message-ID: <1b124954-2804-a64c-e567-0bbf0ae15ed5@gmx.de>
Date: Fri, 19 Jun 2020 05:03:57 +0200
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:68.0) Gecko/20100101 Thunderbird/68.9.0
MIME-Version: 1.0
In-Reply-To: <9B975808-6BF3-491C-8E32-C475824B0E9A@mnot.net>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: quoted-printable
X-Provags-ID: V03:K1:o0QYTXCGbVsvStm1y0oeEVUgLy42Dw3i/rY7eLrP3RYliR3GiSq Cb2umU5QtObrc+yV6Ye5U//TVeO+tc0W4E9qcn4WZ4EW8ctOXUngM5xX5TqQuUaVrBqjBy8 gvdT5TF153QTiEumSp/hO4qdvMICHBBbR4ppI0PHVvb4sy79UaC4KrKI4OvpH6+G+pwE7xP eNaQ9EiWYVWLNJzG5jvUQ==
X-UI-Out-Filterresults: notjunk:1;V03:K0:SN1dVQExIOs=:zWTHcqGSbyD6CsDI6HObSg UoreEQbzempc1mHP4GmaMXgFoG3PDdItKv4OX2S50Q0bZv8sUmiIFavg1m3BLaL1wSGrFleLJ cpeG2zfE1ndavXrT1vgtdZea1Bgh4g7Pj/PL8jCjElGpKMMwGa16uC4yXYyWOYeTHHrPPCIo8 y/2rfxroR4QX86oSBIFiPNTNzzAQVU/4nx3K9Ghi+7/a94tiqcsDfRK4jS4Xk1Q2FcvNlT+/K rL137k6ENRv/Db4Z+8PK5HJ9nPVp+GOgr2oL1YFuHE6jU4OAQnkYtpT34blOskZa2KjxM8lof U7oH1GLqIaCBRSthZXCqwQFW/xrlq2evLn6i1eQ6r+7C3fvY6R7Sqn4vh6nkPjHn6PUEhg6pJ VqrzwupplOfti/dqB0UfUUubizyn4bMMF02iLI9RF24nKrelw9bSp7s7x3C+8j/0bB20VJ6Un NMLy4HmXDYQZxas88fKtJ80w2EO+pDZk3TKjGusxnmlGyGzxICbddEw5muU3FSygzRZvLn3Uh K7JqMvJWXD1MQEceEy8HqzkpThFro6X8R/OmYnnT7r5KxoNHcXH32zacIibdlGeCLPaNBzTog vvnPwH27T62nMmLZXa9HmISFLcVwUK1lfudWxEbMZL+TW08sOXAjFH7XBaXFIoT2l8Fj/5pke geN1XWuEYSidOmG2sY+BsodKc7bZUa5o9SFn4GpzN23dWJHBAxsBCKAUwzm++q/l1L0TycWuy Br/uXdwPNZC67/CRcRjiJJi89O7UJ3fr7XZLnxO97rArVg89LcDH/xBZDnjNkh9erEMWhabMl eO92s7oYchT25zNOkZmb4IuT3PgDUsqEQtMB8hW4JuRzdphQy4dA3O99W1Qzf8hFhAtGg2yCQ BAFO83yHE/l5mI717CECrTJqxPSM2tTi7WfLMliMtapZBE8c2A/8zpc2m0VKt0cOk7wc/0jF9 xfVP+Zk/KG7K/BvFKaGC/j3ngDJBhjqIHuP6iz5awp5wbovVk5a15sfiYCt4/Iw6IKco/Rqz2 CqutTHxJbIIY1GUq2dCG4fZKWOPmB6okTwIuirdcERukGEDpmNr9nlDzKcyRvIIOjtFO1LeBG f/BNhKXbPO4RfIPXtUrmbcMZMGI2Xewuf+Om4kZ8Htq0x+Nxx2bI28kImwQ5jNjeo6em/8o3b /W8e8w66Ge3t6zoY9/eHJMWFn2XAV9Jj7TitQ5Lr0h2k5ktv7Fi0mxzlYJq5wMq3iEQFzSXOb HHEbQP0Txb/Kzv8Os
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/FAgMSSHVjJuD8EDfXoE8b5wJb4A>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 19 Jun 2020 03:04:07 -0000

On 19.06.2020 01:19, Mark Nottingham wrote:
> My .02 -
>
> I think we need to balance between being correct and over-complicating t=
he text here. In *theory*, Julian is correct; in wide practice, pipelining=
 isn't used because of the limitations it has, as well as implementation /=
 interop issues.
>
> How about something like:
>
> "Because of limitations in how HTTP exchanges are mapped onto TCP connec=
tions in HTTP/1.1, multiple parallel connections are often used..."
>
> ?
>
> Cheers,

SGTM.


From nobody Thu Jun 18 21:42:21 2020
Return-Path: <kazuhooku@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id F39EA3A0E00 for <quic@ietfa.amsl.com>; Thu, 18 Jun 2020 21:42:19 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id LALbfnxgJUVu for <quic@ietfa.amsl.com>; Thu, 18 Jun 2020 21:42:18 -0700 (PDT)
Received: from mail-ed1-x534.google.com (mail-ed1-x534.google.com [IPv6:2a00:1450:4864:20::534]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 9C7AF3A0DE3 for <quic@ietf.org>; Thu, 18 Jun 2020 21:42:17 -0700 (PDT)
Received: by mail-ed1-x534.google.com with SMTP id x93so6516579ede.9 for <quic@ietf.org>; Thu, 18 Jun 2020 21:42:17 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=XyBxYXGJF2MhvPxFUCxjOHIendx9tKZK3vHxQq5vRbg=; b=iy9Wa9+j1kd5+B6nFuJvGUZG+3o2H1d20KTIZsNiq8MSL9VRcX1MHE2q1mhsCiGvqN KSX2A18UQKq//kdcJ/MwDkrADQEs587/5I4weyuTb56CTxU8DtYPmQ6+29Yep/vD+CtM sIVC1OBS0q5prm4tUomw24N0bKh4Kw1uRS/f+3S5tYmlXW/77/aEo/TKXL+Y0rESt/W9 cJVR2PfU0n75DUO36vGpy30AAO07ib7MDGlx1vclEXT9iKcvd0NAaDa2Tq0Ay63HZ2VH YsebShaYAqNH5rrkHhvNLYMtM91ZMjduCVtbaiOL8N3wXN63w5BTLbQqu5rKfwT0MObc rRkg==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=XyBxYXGJF2MhvPxFUCxjOHIendx9tKZK3vHxQq5vRbg=; b=FHFMcdXYMo7BL8LKOzpPZe5z+i115WN4yDxOSATLL5GNWUPIpnROlW1j9GSTFQHLxH WFqdIbsdV4gmancCYcYnYExCEj+cmkgTllcTOWVKBj7KYeBD1RCsH/ol/91fk5TNV8Pv c2tyy8gUijbbDLThhzVbFfmTGVOkushT+BpbXays7eDOmMR0lIKfZ5PnuyHnBjmQel0M VfvUzSJKVBoagXfE6LZ41x9QRPynRrP+DC0KyaxeiLeYXJvFu6LoduAcpG4sQAvsMX3S 8T4Zmz3ho7d6BbTYCGx/DEBUp7awr8315x3qydr7o8cKmcNswyAJtGxB7gfXnnPkCxkG XjNg==
X-Gm-Message-State: AOAM530twmbFbNrRorCzDZyp3NPgkcITPN4xRfMrbtojrdT5y4MkJahp x+zLLFckIeevryTRHZEjXJyaMxHaID9892cHilw=
X-Google-Smtp-Source: ABdhPJwMOPvB++2/thkW4YiK7jyvUwWPYyDxnegJYBUhkDzYWbn3osEHB6mqVk3lmcxz0TjCybJyKZApvQMw4m2DrpQ=
X-Received: by 2002:a05:6402:642:: with SMTP id u2mr1452742edx.230.1592541736035;  Thu, 18 Jun 2020 21:42:16 -0700 (PDT)
MIME-Version: 1.0
References: <CANatvzz8F1H=DXMkBEhmKHnYM-HVG48TS9KwY=OP881Txkcodw@mail.gmail.com> <CADdTf+i+LZ98GgNhFNVcuoVczC=jCQE-TqWbCqhrpR7=Z2knWg@mail.gmail.com> <CAN1APdft3UU1dfKY_UxRaLy2xeCYSXQT3=k53=96OO_Gu1X_cw@mail.gmail.com> <CANatvzw2pj-CaXbAJ_-kUrvmoC3_oyvnSo7Yn+mX-kBuqVkr3w@mail.gmail.com> <CADdTf+jtAd5rh+fW2RMLjTgXzNZ30xxSQMXKRtRb4z7bZcRsbg@mail.gmail.com>
In-Reply-To: <CADdTf+jtAd5rh+fW2RMLjTgXzNZ30xxSQMXKRtRb4z7bZcRsbg@mail.gmail.com>
From: Kazuho Oku <kazuhooku@gmail.com>
Date: Fri, 19 Jun 2020 13:42:05 +0900
Message-ID: <CANatvzxsPhZvyLtk80aPCnr+g7rdpDngvSLxp6eWRYppEHjTrQ@mail.gmail.com>
Subject: Re: AEAD and header encryption overhead in QUIC
To: Matt Joras <matt.joras@gmail.com>
Cc: =?UTF-8?Q?Mikkel_Fahn=C3=B8e_J=C3=B8rgensen?= <mikkelfj@gmail.com>,  Nick Banks <nibanks@microsoft.com>, IETF QUIC WG <quic@ietf.org>
Content-Type: multipart/alternative; boundary="00000000000077698f05a868844a"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/VBgI7A6sxzU8UhVmQ9jE7YrpkPI>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 19 Jun 2020 04:42:20 -0000

--00000000000077698f05a868844a
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

2020=E5=B9=B46=E6=9C=8819=E6=97=A5(=E9=87=91) 8:24 Matt Joras <matt.joras@g=
mail.com>:

> Thanks for the explanations, Kazuho! Would it be possible to share the
> commands you used to produce the given results? It would be interesting s=
o
> others can do direct comparisons with their own performance benchmarking.
> It is always hard to do comparisons across implementations with varying
> hardware and OS versions.
>

That's a reasonable thing to ask.

While I'm not fully certain if it is helpful to have direct
comparison between the stacks (everybody has different requirements /
deployments), I am at least certain that sharing the setup being used is a
good idea. It also helps me reproduce the results in the future.

I've written down the setup that we used for the benchmark shown in
https://github.com/h2o/quicly/pull/359 at
https://github.com/h2o/quicly/wiki/Benchmarking-CPU-usage. Regarding the
AES-GCM benchmark, I think you can find sufficient information in the last
paragraph of https://github.com/h2o/picotls/pull/310.


>
> Matt Joras
>
> On Thu, Jun 18, 2020 at 2:40 PM Kazuho Oku <kazuhooku@gmail.com> wrote:
>
>> Nick, Matt, Mikkel, thank you for your comments.
>>
>> 2020=E5=B9=B46=E6=9C=8818=E6=97=A5(=E6=9C=A8) 23:52 Nick Banks <nibanks@=
microsoft.com>:
>>
>>> We=E2=80=99ve found that an easy way to minimize the CPU cost of the he=
ader
>>> protection (on send and receive) is batching. If you copy 8 packet head=
ers
>>> into a single contiguous block of memory, you can do a single crypto
>>> operation on it. The cost of doing one batch of 8 is essentially the sa=
me
>>> as doing just a single header. This effectively cuts the cost of header
>>> protection to 1/8th the original cost. Obviously this only works if you
>>> have enough packets to batch process, but for high throughput tests, yo=
u
>>> usually do.
>>>
>>>
>>>
>>> Feel free to take a look at the MsQuic code (receive path
>>> <https://github.com/microsoft/msquic/blob/master/src/core/connection.c#=
L4784>
>>> , send path
>>> <https://github.com/microsoft/msquic/blob/master/src/core/packet_builde=
r.c#L728>).
>>> The nice thing about this design (for those of us who don=E2=80=99t hav=
e as much
>>> special crypto expertise) is that it doesn=E2=80=99t require any specia=
l crypto
>>> changes. I=E2=80=99d be interested to see if our two approaches could b=
e combined
>>> for even better performance!
>>>
>>
>> Thank you for sharing your experience. This is indeed a sensible approac=
h
>> for minimizing the cost of header protection. On the receive side,
>> processing multiple packets in batch is the only way of minimizing the c=
ost
>> of header protection, because AEAD unprotection cannot start until the
>> result of header unprotection is being obtained. Considering that modern
>> x86-64 CPUs can run 4 to 8 AES block operations in parallel, doing heade=
r
>> protection for one packet at a time is a waste of CPU resources.
>>
>> To paraphrase, the two approaches are orthogonal on the receive-side, an=
d
>> combining them would give better performance. My napkin tells me that I =
can
>> expect at most 3% reduction in CPU cycles spent in crypto.
>>
>> On the send side however, I do not expect to see performance increase by
>> combining the two approaches. This "game" is about keeping the CPU
>> pipelines that do AES-NI running at their full speed. The cost does not
>> change as long as the AES operations of header protection are run in
>> parallel with other AES operations, regardless of that being AEAD or hea=
der
>> protection o other packets.
>>
>> If there's chance of further improving performance, I tend to think that
>> that would come from overlapping the AEAD operation of multiple packets.
>> This "game" is about keeping the AES-NI pipeline busy. With Fusion, we'v=
e
>> reached 90% of the theoretical maximum speed (9th-gen Intel Core CPUs ca=
n
>> do 64 bytes of AES encryption in 40 clocks). The question to us is if we
>> want to consider changing our API for single-digit performance gain.
>>
>>
>> 2020=E5=B9=B46=E6=9C=8819=E6=97=A5(=E9=87=91) 2:21 Matt Joras <matt.jora=
s@gmail.com>:
>>
>>> I was curious, since it's not mentioned on the PRs that I saw, were
>>> these tests done with any adjustments to the ACK frequency on the clien=
t?
>>> Or is this with the default ACK policy of ACKing every other?
>>>
>>
>> This is an important point, thank you for asking. We did reduce ACK
>> frequency to 1/8 CWND, as we did in our previous report [1]. In fact, yo=
u
>> cannot turn that off with quicly.
>>
>>
>> 2020=E5=B9=B46=E6=9C=8819=E6=97=A5(=E9=87=91) 2:35 Mikkel Fahn=C3=B8e J=
=C3=B8rgensen <mikkelfj@gmail.com>:
>>
>>> I am a bit surprised that it is possible to use bulk encrypt headers as
>>> Nick suggests as I would have thought the key / nonce was associated wi=
th
>>> the packet, but I haven=E2=80=99t looked closely recently.
>>>
>>
>> That's a keen observation. Nonce used for header protection does depend
>> on the output of AEAD. Though, because the nonce is taken from the very
>> first few bytes of AEAD output, it is possible to start calculating the
>> header protection vector before finishing AEAD operation, unless the pac=
ket
>> is tiny. The position of the nonce was deliberately chosen to provide ro=
om
>> for this type of optimization. I recall at least one person (Martin
>> Thomson) arguing for having this possibility.
>>
>> [1]
>> https://www.fastly.com/blog/measuring-quic-vs-tcp-computational-efficien=
cy
>>
>> --
>> Kazuho Oku
>>
>

--=20
Kazuho Oku

--00000000000077698f05a868844a
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div dir=3D"ltr"><div dir=3D"ltr"><div dir=3D"ltr"><div di=
r=3D"ltr"><div dir=3D"ltr"><br></div><br><div class=3D"gmail_quote"><div di=
r=3D"ltr" class=3D"gmail_attr">2020=E5=B9=B46=E6=9C=8819=E6=97=A5(=E9=87=91=
) 8:24 Matt Joras &lt;<a href=3D"mailto:matt.joras@gmail.com">matt.joras@gm=
ail.com</a>&gt;:<br></div><blockquote class=3D"gmail_quote" style=3D"margin=
:0px 0px 0px 0.8ex;border-left-width:1px;border-left-style:solid;border-lef=
t-color:rgb(204,204,204);padding-left:1ex"><div dir=3D"ltr"><div>Thanks for=
 the explanations, Kazuho! Would it be possible to share the commands you u=
sed to produce the given results? It would be interesting so others can do =
direct comparisons with their own performance benchmarking. It is always ha=
rd to do comparisons across implementations with varying hardware and OS ve=
rsions.</div></div></blockquote><div><br></div><div>That&#39;s a reasonable=
 thing to ask.</div><div><br></div><div>While I&#39;m not fully certain if =
it is helpful to have direct comparison=C2=A0between the stacks (everybody =
has different requirements / deployments), I am at least certain that shari=
ng the setup being used is a good idea. It also helps me reproduce the resu=
lts in the future.</div><div><br></div><div>I&#39;ve written down the setup=
 that we used for the benchmark shown in <a href=3D"https://github.com/h2o/=
quicly/pull/359">https://github.com/h2o/quicly/pull/359</a>=C2=A0at <a href=
=3D"https://github.com/h2o/quicly/wiki/Benchmarking-CPU-usage">https://gith=
ub.com/h2o/quicly/wiki/Benchmarking-CPU-usage</a>. Regarding the AES-GCM be=
nchmark, I think you can find sufficient information in the last paragraph =
of=C2=A0<a href=3D"https://github.com/h2o/picotls/pull/310">https://github.=
com/h2o/picotls/pull/310</a>.</div><div>=C2=A0</div><blockquote class=3D"gm=
ail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left-width:1px;border-l=
eft-style:solid;border-left-color:rgb(204,204,204);padding-left:1ex"><div d=
ir=3D"ltr"><div><br></div><div>Matt Joras<br></div></div><br><div class=3D"=
gmail_quote"><div dir=3D"ltr" class=3D"gmail_attr">On Thu, Jun 18, 2020 at =
2:40 PM Kazuho Oku &lt;<a href=3D"mailto:kazuhooku@gmail.com" target=3D"_bl=
ank">kazuhooku@gmail.com</a>&gt; wrote:<br></div><blockquote class=3D"gmail=
_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left=
-style:solid;border-left-color:rgb(204,204,204);padding-left:1ex"><div dir=
=3D"ltr"><div dir=3D"ltr"><div dir=3D"ltr"><div dir=3D"ltr"><div dir=3D"ltr=
"><div dir=3D"ltr"><div dir=3D"ltr"><div dir=3D"ltr">Nick, Matt, Mikkel, th=
ank you for your comments.<div><br></div><div><span style=3D"color:rgb(0,0,=
0)">2020=E5=B9=B46=E6=9C=8818=E6=97=A5(=E6=9C=A8) 23:52 Nick Banks &lt;<a h=
ref=3D"mailto:nibanks@microsoft.com" target=3D"_blank">nibanks@microsoft.co=
m</a>&gt;:</span><br></div><div><blockquote class=3D"gmail_quote" style=3D"=
color:rgb(0,0,0);margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left=
-style:solid;border-left-color:rgb(204,204,204);padding-left:1ex"><div lang=
=3D"EN-US"><div><p class=3D"MsoNormal">We=E2=80=99ve found that an easy way=
 to minimize the CPU cost of the header protection (on send and receive) is=
 batching. If you copy 8 packet headers into a single contiguous block of m=
emory, you can do a single crypto operation on it. The cost of doing one ba=
tch of 8 is essentially the same as doing just a single header. This effect=
ively cuts the cost of header protection to 1/8<sup>th</sup>=C2=A0the origi=
nal cost. Obviously this only works if you have enough packets to batch pro=
cess, but for high throughput tests, you usually do.<br></p><p class=3D"Mso=
Normal"><u></u>=C2=A0<u></u></p><p class=3D"MsoNormal">Feel free to take a =
look at the MsQuic code (<a href=3D"https://github.com/microsoft/msquic/blo=
b/master/src/core/connection.c#L4784" target=3D"_blank">receive path</a>,=
=C2=A0<a href=3D"https://github.com/microsoft/msquic/blob/master/src/core/p=
acket_builder.c#L728" target=3D"_blank">send path</a>). The nice thing abou=
t this design (for those of us who don=E2=80=99t have as much special crypt=
o expertise) is that it doesn=E2=80=99t require any special crypto changes.=
 I=E2=80=99d be interested to see if our two approaches could be combined f=
or even better performance!</p></div></div></blockquote><div><br></div><div=
>Thank you for sharing your experience. This is indeed a sensible approach =
for minimizing the cost of header protection. On the receive side, processi=
ng multiple packets in batch is the only way of minimizing the cost of head=
er protection, because AEAD unprotection cannot start until the result of h=
eader unprotection is being obtained. Considering that modern x86-64 CPUs c=
an run 4 to 8 AES block operations in parallel, doing header protection for=
 one packet at a time is a waste of CPU resources.</div><div><br></div><div=
>To paraphrase, the two approaches are orthogonal on the receive-side, and =
combining them would give better performance. My napkin tells me that I can=
 expect at most 3% reduction in CPU cycles spent in crypto.</div><div><br><=
/div><div>On the send side however, I do not expect to see performance incr=
ease by combining the two approaches. This &quot;game&quot; is about keepin=
g the CPU pipelines that do AES-NI running at their full speed. The cost do=
es not change as long as the AES operations of header protection are run in=
 parallel with other AES operations, regardless of that being AEAD or heade=
r protection o other packets.</div><div><br></div><div>If there&#39;s chanc=
e of further improving performance, I tend to think that that would come fr=
om overlapping the AEAD operation of multiple packets. This &quot;game&quot=
; is about keeping the AES-NI pipeline busy. With Fusion, we&#39;ve reached=
 90% of the theoretical maximum speed (9th-gen Intel Core CPUs can do 64 by=
tes of AES encryption in 40 clocks). The question to us is if we want to co=
nsider changing our API for single-digit performance gain.</div><div>=C2=A0=
</div></div><div><br></div><div><div dir=3D"ltr" class=3D"gmail_attr" style=
=3D"color:rgb(0,0,0)">2020=E5=B9=B46=E6=9C=8819=E6=97=A5(=E9=87=91) 2:21 Ma=
tt Joras &lt;<a href=3D"mailto:matt.joras@gmail.com" target=3D"_blank">matt=
.joras@gmail.com</a>&gt;:</div><blockquote class=3D"gmail_quote" style=3D"c=
olor:rgb(0,0,0);margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-=
style:solid;border-left-color:rgb(204,204,204);padding-left:1ex"><div dir=
=3D"ltr">I was curious, since it&#39;s not mentioned on the PRs that I saw,=
 were these tests done with any adjustments to the ACK frequency on the cli=
ent? Or is this with the default ACK policy of ACKing every other?</div></b=
lockquote><div><br></div><div>This is an important point, thank you for ask=
ing. We did reduce ACK frequency to 1/8 CWND, as we did in our previous rep=
ort [1]. In fact, you cannot turn that off with quicly.</div></div><div><br=
></div><div><br></div></div></div></div></div></div></div><div class=3D"gma=
il_quote"><div dir=3D"ltr" class=3D"gmail_attr">2020=E5=B9=B46=E6=9C=8819=
=E6=97=A5(=E9=87=91) 2:35 Mikkel Fahn=C3=B8e J=C3=B8rgensen &lt;<a href=3D"=
mailto:mikkelfj@gmail.com" target=3D"_blank">mikkelfj@gmail.com</a>&gt;:</d=
iv><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;bord=
er-left-width:1px;border-left-style:solid;border-left-color:rgb(204,204,204=
);padding-left:1ex"><div><div style=3D"font-family:Helvetica,Arial;font-siz=
e:13px">I am a bit surprised that it is possible to use bulk encrypt header=
s as Nick suggests as I would have thought the key / nonce was associated w=
ith the packet, but I haven=E2=80=99t looked closely recently.</div> </div>=
</blockquote></div><div><br></div><div>That&#39;s a keen observation. Nonce=
 used for header protection does depend on the output of AEAD. Though, beca=
use the nonce is taken from the very first few bytes of AEAD output, it is =
possible to start calculating the header protection vector before finishing=
 AEAD operation, unless the packet is tiny. The position of the nonce was d=
eliberately chosen to provide room for this type of optimization. I recall =
at least one person (Martin Thomson) arguing for having this possibility.</=
div><div><br></div><div>[1]=C2=A0<a href=3D"https://www.fastly.com/blog/mea=
suring-quic-vs-tcp-computational-efficiency" target=3D"_blank">https://www.=
fastly.com/blog/measuring-quic-vs-tcp-computational-efficiency</a></div><di=
v><br></div>-- <br><div dir=3D"ltr">Kazuho Oku</div></div></div>
</blockquote></div>
</blockquote></div><br clear=3D"all"><div><br></div>-- <br><div dir=3D"ltr"=
 class=3D"gmail_signature">Kazuho Oku</div></div></div></div></div></div>

--00000000000077698f05a868844a--


From nobody Fri Jun 19 08:08:26 2020
Return-Path: <ilubashe@akamai.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 75FD13A0AC2 for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 08:08:24 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.099
X-Spam-Level: 
X-Spam-Status: No, score=-2.099 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=unavailable autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=akamai.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id f0z9F9FIiQgz for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 08:08:20 -0700 (PDT)
Received: from mx0b-00190b01.pphosted.com (mx0b-00190b01.pphosted.com [IPv6:2620:100:9005:57f::1]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id AE34D3A0AB1 for <quic@ietf.org>; Fri, 19 Jun 2020 08:08:20 -0700 (PDT)
Received: from pps.filterd (m0122331.ppops.net [127.0.0.1]) by mx0b-00190b01.pphosted.com (8.16.0.42/8.16.0.42) with SMTP id 05JExAaE020891; Fri, 19 Jun 2020 16:08:10 +0100
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=akamai.com; h=from : to : cc : subject : date : message-id : references : in-reply-to : content-type : mime-version; s=jan2016.eng; bh=1iYul8mYgNQlzVVcJ3LIpz6OOnDA5HEzWEUNlrAd/AU=; b=S9veIkkSB6QL2bmV/PCek4ZO/vsVvMK8PAJBxExhR3q/G77OXqeeIFpUHqORFhKHefvn zKW7FCrX+MpVKXv1SU4/EQ5oyyalp0KTF7X0AKNLoZ77E5ZPdKW8SDManqWq3+7ADwYI 6dG5NFhXn8F70/XczFz31tfv5Oq0P/BooUdyp7TzHfB0F7ydoLbjyPgaFXJAMuVR9XJe N5XJQX3FXc6A9Wzv1aoJx+q3tGlvn7DgRaGBJcTQxfL4sbfC6lOaFVg9H7snqAktuCOy GHE13vBotnN5MeeRijFDDVjmatACzV68TOXxUm4XHal7HcRWTzZJLTKWvDKXenSr/u9w gw== 
Received: from prod-mail-ppoint1 (prod-mail-ppoint1.akamai.com [184.51.33.18] (may be forged)) by mx0b-00190b01.pphosted.com with ESMTP id 31qrdxq4yw-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Fri, 19 Jun 2020 16:08:10 +0100
Received: from pps.filterd (prod-mail-ppoint1.akamai.com [127.0.0.1]) by prod-mail-ppoint1.akamai.com (8.16.0.42/8.16.0.42) with SMTP id 05JEnq7R025266; Fri, 19 Jun 2020 11:08:09 -0400
Received: from email.msg.corp.akamai.com ([172.27.123.34]) by prod-mail-ppoint1.akamai.com with ESMTP id 31qjmbvger-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-SHA384 bits=256 verify=NOT); Fri, 19 Jun 2020 11:08:08 -0400
Received: from USMA1EX-DAG1MB5.msg.corp.akamai.com (172.27.123.105) by usma1ex-dag1mb2.msg.corp.akamai.com (172.27.123.102) with Microsoft SMTP Server (TLS) id 15.0.1497.2; Fri, 19 Jun 2020 11:08:08 -0400
Received: from USMA1EX-DAG1MB5.msg.corp.akamai.com ([172.27.123.105]) by usma1ex-dag1mb5.msg.corp.akamai.com ([172.27.123.105]) with mapi id 15.00.1497.006; Fri, 19 Jun 2020 11:08:08 -0400
From: "Lubashev, Igor" <ilubashe@akamai.com>
To: Martin Duke <martin.h.duke@gmail.com>
CC: Christian Huitema <huitema@huitema.net>, Kyle Rose <krose@krose.org>, "Ian Swett" <ianswett=40google.com@dmarc.ietf.org>, Lars Eggert <lars@eggert.org>, IETF QUIC WG <quic@ietf.org>, Jared Mauch <jared@puck.nether.net>
Subject: RE: Is the invariants draft really standards track?
Thread-Topic: Is the invariants draft really standards track?
Thread-Index: AQHWM4bg7m/IhU9tKEaMc9qs7omEcqi7wjeAgACBhgCAAA8vAIAAAZ8A///Fg1CAId32AIACNYSA
Date: Fri, 19 Jun 2020 15:08:07 +0000
Message-ID: <95dd02c92b32472d9cab0dd47b98c637@usma1ex-dag1mb5.msg.corp.akamai.com>
References: <CAM4esxQBqfrz24riPQA_VGKcGp_TzW0pqb97KfFMtNdW9pUfDg@mail.gmail.com> <833A693C-62E6-4889-9954-FCE65A839A7C@eggert.org> <CAKcm_gPMO2DtqvKucqVw0zDjSniSOmFD4p1Tp4YLjr9WSWdEUw@mail.gmail.com> <CAJU8_nUN42gGmQof24XD9-EjXedyzcarDyRP8MGe1qW-BZ=+Aw@mail.gmail.com> <9cd91c24-c730-22a4-7aa0-baf61613b3ce@huitema.net> <f4922cdb59014202900de44cc5fea0ff@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQvwkTvpUcu6-+W5zWo22m-R1jvN7DcCpXfuw8Hb55qsw@mail.gmail.com>
In-Reply-To: <CAM4esxQvwkTvpUcu6-+W5zWo22m-R1jvN7DcCpXfuw8Hb55qsw@mail.gmail.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
x-ms-exchange-transport-fromentityheader: Hosted
x-originating-ip: [172.19.39.208]
Content-Type: multipart/alternative; boundary="_000_95dd02c92b32472d9cab0dd47b98c637usma1exdag1mb5msgcorpak_"
MIME-Version: 1.0
X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.216, 18.0.687 definitions=2020-06-19_16:2020-06-19, 2020-06-19 signatures=0
X-Proofpoint-Spam-Details: rule=notspam policy=default score=0 spamscore=0 malwarescore=0 bulkscore=0 adultscore=0 mlxlogscore=999 mlxscore=0 phishscore=0 suspectscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2004280000 definitions=main-2006190106
X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.216, 18.0.687 definitions=2020-06-19_16:2020-06-19, 2020-06-19 signatures=0
X-Proofpoint-Spam-Details: rule=notspam policy=default score=0 priorityscore=1501 spamscore=0 clxscore=1011 lowpriorityscore=0 malwarescore=0 mlxlogscore=999 impostorscore=0 bulkscore=0 cotscore=-2147483648 phishscore=0 adultscore=0 mlxscore=0 suspectscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2004280000 definitions=main-2006190111
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/izg_GnAg72CPBczbeV5uMiqjiNw>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 19 Jun 2020 15:08:24 -0000

--_000_95dd02c92b32472d9cab0dd47b98c637usma1exdag1mb5msgcorpak_
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_95dd02c92b32472d9cab0dd47b98c637usma1exdag1mb5msgcorpak_
Content-Type: text/html; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_95dd02c92b32472d9cab0dd47b98c637usma1exdag1mb5msgcorpak_--


From nobody Fri Jun 19 10:33:27 2020
Return-Path: <fielding@gbiv.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 5D3823A0D06 for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 10:33:26 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.1
X-Spam-Level: 
X-Spam-Status: No, score=-2.1 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_MSPIKE_H2=-0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=gbiv.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 4aez6I8PoldU for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 10:33:25 -0700 (PDT)
Received: from antelope.elm.relay.mailchannels.net (antelope.elm.relay.mailchannels.net [23.83.212.4]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 1975F3A0D00 for <quic@ietf.org>; Fri, 19 Jun 2020 10:33:24 -0700 (PDT)
X-Sender-Id: dreamhost|x-authsender|fielding@gbiv.com
Received: from relay.mailchannels.net (localhost [127.0.0.1]) by relay.mailchannels.net (Postfix) with ESMTP id 322333603C0; Fri, 19 Jun 2020 17:33:24 +0000 (UTC)
Received: from pdx1-sub0-mail-a25.g.dreamhost.com (100-96-137-10.trex.outbound.svc.cluster.local [100.96.137.10]) (Authenticated sender: dreamhost) by relay.mailchannels.net (Postfix) with ESMTPA id 4CA9F361534; Fri, 19 Jun 2020 17:33:23 +0000 (UTC)
X-Sender-Id: dreamhost|x-authsender|fielding@gbiv.com
Received: from pdx1-sub0-mail-a25.g.dreamhost.com (pop.dreamhost.com [64.90.62.162]) (using TLSv1.2 with cipher DHE-RSA-AES256-GCM-SHA384) by 0.0.0.0:2500 (trex/5.18.8); Fri, 19 Jun 2020 17:33:24 +0000
X-MC-Relay: Neutral
X-MailChannels-SenderId: dreamhost|x-authsender|fielding@gbiv.com
X-MailChannels-Auth-Id: dreamhost
X-Hysterical-Cold: 191588080c4c8c73_1592588003773_3033306100
X-MC-Loop-Signature: 1592588003772:3879789484
X-MC-Ingress-Time: 1592588003772
Received: from pdx1-sub0-mail-a25.g.dreamhost.com (localhost [127.0.0.1]) by pdx1-sub0-mail-a25.g.dreamhost.com (Postfix) with ESMTP id F2DD6ADFE2; Fri, 19 Jun 2020 10:33:22 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha1; c=relaxed; d=gbiv.com; h=content-type :mime-version:subject:from:in-reply-to:date:cc :content-transfer-encoding:message-id:references:to; s=gbiv.com; bh=k5W+IMIKAJRKatSvuniFITm16L4=; b=VGCDIW86wZDJWjsznv8wlZVqDzC6 Uuv6yI4SdOP598ThBWNKYrV07PTQDL31Ietq3f0lW/QCF0UoDvMzvDoocffvd91z 5SNsynYE4u+u9L1fZZ5hezC9UvYhm+BdBmWn1am+JHfX+drPB79rw2bY9Z/u+E1u jKzwSmekiIkJWpA=
Received: from [192.168.1.11] (ip68-228-81-25.oc.oc.cox.net [68.228.81.25]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) (Authenticated sender: fielding@gbiv.com) by pdx1-sub0-mail-a25.g.dreamhost.com (Postfix) with ESMTPSA id 7AC1CADFF2; Fri, 19 Jun 2020 10:33:21 -0700 (PDT)
Content-Type: text/plain; charset=us-ascii
Mime-Version: 1.0 (Mac OS X Mail 12.4 \(3445.104.14\))
Subject: Re: draft-ietf-quic-http-29, "1.1. Prior versions of HTTP"
X-DH-BACKEND: pdx1-sub0-mail-a25
From: "Roy T. Fielding" <fielding@gbiv.com>
In-Reply-To: <9B975808-6BF3-491C-8E32-C475824B0E9A@mnot.net>
Date: Fri, 19 Jun 2020 10:33:19 -0700
Cc: Julian Reschke <julian.reschke@gmx.de>, quic@ietf.org
Content-Transfer-Encoding: quoted-printable
Message-Id: <7DD3EE98-199B-4F09-A8C8-1CE604989136@gbiv.com>
References: <8c747b59-1672-4ccb-dfce-0914e74c504f@gmx.de> <9B975808-6BF3-491C-8E32-C475824B0E9A@mnot.net>
To: Mark Nottingham <mnot@mnot.net>
X-Mailer: Apple Mail (2.3445.104.14)
X-VR-OUT-STATUS: OK
X-VR-OUT-SCORE: -100
X-VR-OUT-SPAMCAUSE: gggruggvucftvghtrhhoucdtuddrgeduhedrudejiedguddugecutefuodetggdotefrodftvfcurfhrohhfihhlvgemucggtfgfnhhsuhgsshgtrhhisggvpdfftffgtefojffquffvnecuuegrihhlohhuthemuceftddtnecusecvtfgvtghiphhivghnthhsucdlqddutddtmdenucfjughrpegtggfuhfgjfffgkfhfvffosehtqhhmtdhhtddvnecuhfhrohhmpedftfhohicuvfdrucfhihgvlhguihhnghdfuceofhhivghlughinhhgsehgsghivhdrtghomheqnecuggftrfgrthhtvghrnhepheefjedtheejffekffelffekveeuvdetteffheffvdfgteefudegtdevteduuddunecukfhppeeikedrvddvkedrkedurddvheenucevlhhushhtvghrufhiiigvpedtnecurfgrrhgrmhepmhhouggvpehsmhhtphdphhgvlhhopegludelvddrudeikedruddruddungdpihhnvghtpeeikedrvddvkedrkedurddvhedprhgvthhurhhnqdhprghthhepfdftohihucfvrdcuhfhivghlughinhhgfdcuoehfihgvlhguihhnghesghgsihhvrdgtohhmqedpmhgrihhlfhhrohhmpehfihgvlhguihhnghesghgsihhvrdgtohhmpdhnrhgtphhtthhopehquhhitgesihgvthhfrdhorhhg
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/o6eYvfH_aqGYTHb4-MZAQDd7_tg>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 19 Jun 2020 17:33:26 -0000

> On Jun 18, 2020, at 4:19 PM, Mark Nottingham <mnot@mnot.net> wrote:
>=20
> My .02 -
>=20
> I think we need to balance between being correct and over-complicating =
the text here. In *theory*, Julian is correct; in wide practice, =
pipelining isn't used because of the limitations it has, as well as =
implementation / interop issues.

No, it isn't used by browsers because they aren't aware of the purpose
for each connection and request. API clients almost always use =
pipelines.

> How about something like:
>=20
> "Because of limitations in how HTTP exchanges are mapped onto TCP =
connections in HTTP/1.1, multiple parallel connections are often =
used..."

Why don't you just say

  "Because HTTP/1.x does not include a multiplexing layer, multiple
   TCP connections are often used to service requests in parallel.
   However, that has a negative impact on congestion control and network
   efficiency, since TCP does not share congestion control across =
multiple
   connections."

because those are the actual technical limitations.

....Roy


From nobody Fri Jun 19 11:44:19 2020
Return-Path: <martin.h.duke@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 799963A0D8B for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 11:44:06 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id jYWfit6w01FN for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 11:44:04 -0700 (PDT)
Received: from mail-il1-x12c.google.com (mail-il1-x12c.google.com [IPv6:2607:f8b0:4864:20::12c]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 5C9B03A0D90 for <quic@ietf.org>; Fri, 19 Jun 2020 11:44:04 -0700 (PDT)
Received: by mail-il1-x12c.google.com with SMTP id z2so10235115ilq.0 for <quic@ietf.org>; Fri, 19 Jun 2020 11:44:04 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=aCZ+8jk9ibscBIzx+sIIo2lrAlyCfv4nnjA8XaXchzs=; b=Cc6aQpD5cfbEYdO10p7pRH65KA/4J7oELjE+gCd2d/houCNJ/VLAInD5iam+uvgvmI BzzZ+kI4IQ96SrxSQkckyArNHQK2jMK1PT14GEUlCmZmBV18MQpQaf82RPkrNcPtowpK hlN/Z/5mkOG1qpo4BmJl7qQwo4u15GPAhgYL34JG470hT/P8BlB5S1qmcerQci6wJG3h GYYmNhXkv8mP6D03kQDOiJre+MOeNauGyw+IyjBL9reXm74WnqvJFA4siD6IzCZRsu6c I3Hq+qGS9iSkVYE2Cy1QVNxKvGedQg1lsPRaMTXeqGZhJ3mDRXe0BRQdu51UFSzbU7Cd NBJg==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=aCZ+8jk9ibscBIzx+sIIo2lrAlyCfv4nnjA8XaXchzs=; b=Xf7TsS6GpE70mQcaJ0+/qKZigXWsPTUPawsXTjXx4b+HhMqQaeDyrU5xvTi2/W0mDS WSfcCFNvl5994jQydBwnvRh5Nh8jF1p4ep+ZmsdWFCNejRmpnB27yTm1S5UAWtiFiPSM DCNaP6jaHfFvw1RfVRJTIop7LvZ3gmiY32TnWlJtRYhRZtr96AQVLeVAemnmNBr5Jzor dUSlQ/lts6GofHZIThT/j589jS7l7NiGqezpePIcag+k0TtNrDYbqdt0xSSl3TZCF3Zj aff/irquZXCfRicfB9qAasNbFdNtH7nGErAe9rBAbhwRXDhG3rUercIU0Dct56llkZ20 gyfA==
X-Gm-Message-State: AOAM533x9IGYopzWuJNRludeisgP5h48v4XIobDc/FabXPETJ8QH4wy/ X3zqy6q4AYCGhZnUOu7oivK9PYvf7S9lw1EzpGI=
X-Google-Smtp-Source: ABdhPJwPlvUhTCDmnu37WiYBQdSwqswYmlgrDVyh4gQWDJBP1K4I0q8Klmucon1M0pMnvxwpSJNu+0a4sV5n4YJli+I=
X-Received: by 2002:a92:de48:: with SMTP id e8mr4839987ilr.249.1592592243529;  Fri, 19 Jun 2020 11:44:03 -0700 (PDT)
MIME-Version: 1.0
References: <CAM4esxQBqfrz24riPQA_VGKcGp_TzW0pqb97KfFMtNdW9pUfDg@mail.gmail.com> <833A693C-62E6-4889-9954-FCE65A839A7C@eggert.org> <CAKcm_gPMO2DtqvKucqVw0zDjSniSOmFD4p1Tp4YLjr9WSWdEUw@mail.gmail.com> <CAJU8_nUN42gGmQof24XD9-EjXedyzcarDyRP8MGe1qW-BZ=+Aw@mail.gmail.com> <9cd91c24-c730-22a4-7aa0-baf61613b3ce@huitema.net> <f4922cdb59014202900de44cc5fea0ff@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQvwkTvpUcu6-+W5zWo22m-R1jvN7DcCpXfuw8Hb55qsw@mail.gmail.com> <95dd02c92b32472d9cab0dd47b98c637@usma1ex-dag1mb5.msg.corp.akamai.com>
In-Reply-To: <95dd02c92b32472d9cab0dd47b98c637@usma1ex-dag1mb5.msg.corp.akamai.com>
From: Martin Duke <martin.h.duke@gmail.com>
Date: Fri, 19 Jun 2020 11:43:52 -0700
Message-ID: <CAM4esxQxxXn27rZEY75-AsHD5VF0fqiV1VDyeSrzQ=-sM7JNCg@mail.gmail.com>
Subject: Re: Is the invariants draft really standards track?
To: "Lubashev, Igor" <ilubashe@akamai.com>
Cc: Christian Huitema <huitema@huitema.net>, Kyle Rose <krose@krose.org>,  Ian Swett <ianswett=40google.com@dmarc.ietf.org>, Lars Eggert <lars@eggert.org>, IETF QUIC WG <quic@ietf.org>, Jared Mauch <jared@puck.nether.net>
Content-Type: multipart/alternative; boundary="000000000000f29bcc05a874461c"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/PhK8bDEte2sX_8zyJQCYAH0qImg>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 19 Jun 2020 18:44:17 -0000

--000000000000f29bcc05a874461c
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

>Unfortunately, Retry system protects only server's memory state and some
CPU cycles spent on crypto.  (Servers still need to decrypt CID to decide
it is invalid, and if the attacker is clever enough to establish one valid
connection and use that CID in a flood, the server will also be decrypting
packets.)

There is no need for servers to decrypt CIDs in QUIC-LB. Presumably the
server has a lookup table for its CIDs.

It is true that Retry Services (and indeed, the Retry concept as a whole)
does nothing to protect network capacity.

On Fri, Jun 19, 2020 at 8:08 AM Lubashev, Igor <ilubashe@akamai.com> wrote:

> It looks like
> https://tools.ietf.org/html/draft-ietf-quic-load-balancers-02#section-5
> is an excellent discussion of Retry mechanics.  It definitely deserves a
> reference from this manageability draft.
>
>
>
> The Retry mechanisms described in LB draft are all cooperating boxes, and
> servers must be aware of them.  Unfortunately, Retry system protects only
> server's memory state and some CPU cycles spent on crypto.  (Servers stil=
l
> need to decrypt CID to decide it is invalid, and if the attacker is cleve=
r
> enough to establish one valid connection and use that CID in a flood, the
> server will also be decrypting packets.)  Retry does nothing to protect
> network resources.
>
>
>
> The PR I opened (https://github.com/quicwg/ops-drafts/pull/94) is about
> uncooperating devices that try to mitigate volumetric network attacks.
>
>
>
>
>
> *From:* Martin Duke <martin.h.duke@gmail.com>
> *Sent:* Wednesday, June 17, 2020 8:16 PM
>
> Hi Igor, you might want to check out the "Retry Services" bit of the
> QUIC-LB draft. This has something to do with the DDoS use case you discus=
s.
>
>
>
> On Wed, May 27, 2020 at 9:07 AM Lubashev, Igor <ilubashe@akamai.com>
> wrote:
>
> I=E2=80=99m working on a manageability draft PR for this (how to rate lim=
it UDP to
> reduce disruption to QUIC if you have to rate limit UDP).  ETA end of the
> week (if I do not get pulled into something again).
>
>
>
> The relevant observation is that DDoS with UDP that is indistinguishable
> from QUIC will happen.  UDP is already the most prevalent DDoS vector,
> since it is easy for a compromised non-admin app to send a flood of huge
> UDP packets (with TCP you get throttled by the congestion controller).  S=
o
> there WILL be DDoS protection devices out there to try to mitigate the
> problem, possibly by observing both directions of the flow and deciding
> whether a packet belongs to a valid flow or not.
>
>
>
> Since such middle boxes will be created, the more explicit and normative
> Invariants are about what one can expect, the less such middle boxes may
> decide for themselves.  For example (I did not think long about it), if
> some elements of path validation could land into Invariants (roughly, =E2=
=80=9Cno
> more than X packets/bytes can be sent on a new path w/o a return packet=
=E2=80=9D),
> a DDoS middle box may use this fact and active connection migration might
> still have a chance during an attack (NAT rebinding could be linked by DD=
oS
> boxes to an old connection via unchanged CID).
>
>
>
>    - Igor
>
>
>
>
>
> *From:* Christian Huitema <huitema@huitema.net>
> *Sent:* Wednesday, May 27, 2020 11:34 AM
>
> On 5/27/2020 8:28 AM, Kyle Rose wrote:
>
> On Wed, May 27, 2020 at 10:34 AM Ian Swett <ianswett=3D
> 40google.com@dmarc.ietf.org> wrote:
>
> I was agreeing with MT, but I'm happy to see some more MUSTs added if
> people feel that'd be helpful.
>
>
>
> Coincidentally, we were just talking about this internally at Akamai
> yesterday. IMO, an invariants document isn't really helpful if it isn't
> normative, and for it to be normative it (or a related practices doc for
> operators) really needs to spell out clear boundaries for operators with
> MUSTs..
>
>
>
> The example that came up yesterday was around operators filtering QUIC in
> the event of a DDoS: one recommendation based on some conversations going
> back at least to Prague 2019 was to hash packets on 4-tuple and filter
> those below a hash value chosen for a desired ingress limit instead of
> doing what most operators do with UDP today, which is to cap UDP throughp=
ut
> and just drop packets randomly or tail drop.
>
> Interesting. Did they consider using the CID, or a fraction of it? This
> looks entirely like the scenario for which we developed stateless retry.
>
>
>
> This recommendation certainly imposes some constraints on future protocol
> development that motivate new invariants: for instance, it would preclude
> sharding a connection across multiple source ports (not that there is
> necessarily a reason to do this; it's just an example). But more
> importantly, it goes beyond invariants: it's one among many practices
> compatible with the current set of invariants, some reasonable and some
> terrible.
>
> This would break the "preferred address" redirection. Preferred address
> migration may or may not be spelled out in the invariants.
>
>
>
> Operators are going to do things to QUIC traffic, so it would be good to
> offer them recommendations that are compatible with broad deployability.
>
>
>
> Yes, we do need the invariants for that.
>
> -- Christian Huitema
>
>

--000000000000f29bcc05a874461c
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">&gt;Unfortunately, Retry system protects only server&#39;s=
 memory state and some CPU cycles spent on crypto.=C2=A0 (Servers still nee=
d to decrypt CID to decide it is invalid, and if the attacker is clever eno=
ugh to establish one valid connection and use that CID in a flood, the serv=
er will also be decrypting packets.)=C2=A0=C2=A0<br><div><br></div><div>The=
re is no need for servers to decrypt CIDs in QUIC-LB. Presumably the server=
 has a lookup table for its CIDs.</div><div><br></div><div>It is true that =
Retry Services (and indeed, the Retry concept as a whole) does nothing to p=
rotect network capacity.</div></div><br><div class=3D"gmail_quote"><div dir=
=3D"ltr" class=3D"gmail_attr">On Fri, Jun 19, 2020 at 8:08 AM Lubashev, Igo=
r &lt;<a href=3D"mailto:ilubashe@akamai.com">ilubashe@akamai.com</a>&gt; wr=
ote:<br></div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px=
 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">





<div lang=3D"EN-US">
<div class=3D"gmail-m_-8614907429145891794WordSection1">
<p class=3D"MsoNormal">It looks like <a href=3D"https://tools.ietf.org/html=
/draft-ietf-quic-load-balancers-02#section-5" target=3D"_blank">https://too=
ls.ietf.org/html/draft-ietf-quic-load-balancers-02#section-5</a> is an exce=
llent discussion of Retry mechanics.=C2=A0 It definitely deserves a referen=
ce from this manageability draft.<u></u><u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal">The Retry mechanisms described in LB draft are all c=
ooperating boxes, and servers must be aware of them.=C2=A0 Unfortunately, R=
etry system protects only server&#39;s memory state and some CPU cycles spe=
nt on crypto.=C2=A0 (Servers still need to decrypt
 CID to decide it is invalid, and if the attacker is clever enough to estab=
lish one valid connection and use that CID in a flood, the server will also=
 be decrypting packets.)=C2=A0 Retry does nothing to protect network resour=
ces.<u></u><u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal">The PR I opened (<a href=3D"https://github.com/quicw=
g/ops-drafts/pull/94" target=3D"_blank">https://github.com/quicwg/ops-draft=
s/pull/94</a>) is about uncooperating devices that try to mitigate volumetr=
ic network attacks.<u></u><u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<div style=3D"border-top:none;border-right:none;border-bottom:none;border-l=
eft:1.5pt solid blue;padding:0in 0in 0in 4pt">
<div>
<div style=3D"border-right:none;border-bottom:none;border-left:none;border-=
top:1pt solid rgb(225,225,225);padding:3pt 0in 0in">
<p class=3D"MsoNormal"><b>From:</b> Martin Duke &lt;<a href=3D"mailto:marti=
n.h.duke@gmail.com" target=3D"_blank">martin.h.duke@gmail.com</a>&gt; <br>
<b>Sent:</b> Wednesday, June 17, 2020 8:16 PM<br>
<br>
<u></u><u></u></p>
</div>
</div>
<div>
<p class=3D"MsoNormal">Hi Igor, you might want to check out the &quot;Retry=
 Services&quot; bit of the QUIC-LB draft. This has something to do with the=
 DDoS use case you discuss.<u></u><u></u></p>
</div>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<div>
<div>
<p class=3D"MsoNormal">On Wed, May 27, 2020 at 9:07 AM Lubashev, Igor &lt;<=
a href=3D"mailto:ilubashe@akamai.com" target=3D"_blank">ilubashe@akamai.com=
</a>&gt; wrote:<u></u><u></u></p>
</div>
<blockquote style=3D"border-top:none;border-right:none;border-bottom:none;b=
order-left:1pt solid rgb(204,204,204);padding:0in 0in 0in 6pt;margin-left:4=
.8pt;margin-right:0in">
<div>
<div>
<p class=3D"MsoNormal">I=E2=80=99m working on a manageability draft PR for =
this (how to rate limit UDP to reduce disruption to QUIC if you have to rat=
e limit UDP).=C2=A0 ETA end of the week (if I do not get pulled
 into something again).<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<p class=3D"MsoNormal">The relevant observation is that DDoS with UDP that =
is indistinguishable from QUIC will happen.=C2=A0 UDP is already the most p=
revalent DDoS vector, since it is easy for a compromised
 non-admin app to send a flood of huge UDP packets (with TCP you get thrott=
led by the congestion controller).=C2=A0 So there WILL be DDoS protection d=
evices out there to try to mitigate the problem, possibly by observing both=
 directions of the flow and deciding
 whether a packet belongs to a valid flow or not.<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<p class=3D"MsoNormal">Since such middle boxes will be created, the more ex=
plicit and normative Invariants are about what one can expect, the less suc=
h middle boxes may decide for themselves.=C2=A0 For example
 (I did not think long about it), if some elements of path validation could=
 land into Invariants (roughly, =E2=80=9Cno more than X packets/bytes can b=
e sent on a new path w/o a return packet=E2=80=9D), a DDoS middle box may u=
se this fact and active connection migration might
 still have a chance during an attack (NAT rebinding could be linked by DDo=
S boxes to an old connection via unchanged CID).<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<ul type=3D"disc">
<li class=3D"MsoNormal">
Igor<u></u><u></u></li></ul>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<div style=3D"border-top:none;border-right:none;border-bottom:none;border-l=
eft:1.5pt solid blue;padding:0in 0in 0in 4pt">
<p class=3D"MsoNormal" style=3D"margin-bottom:12pt"><b>From:</b> Christian =
Huitema &lt;<a href=3D"mailto:huitema@huitema.net" target=3D"_blank">huitem=
a@huitema.net</a>&gt;
<br>
<b>Sent:</b> Wednesday, May 27, 2020 11:34 AM<u></u><u></u></p>
<div>
<p class=3D"MsoNormal">On 5/27/2020 8:28 AM, Kyle Rose wrote:<u></u><u></u>=
</p>
</div>
<blockquote style=3D"margin-top:5pt;margin-bottom:5pt">
<p class=3D"MsoNormal">On Wed, May 27, 2020 at 10:34 AM Ian Swett &lt;iansw=
ett=3D<a href=3D"mailto:40google.com@dmarc.ietf.org" target=3D"_blank">40go=
ogle.com@dmarc.ietf.org</a>&gt; wrote:<u></u><u></u></p>
<div>
<blockquote style=3D"border-top:none;border-right:none;border-bottom:none;b=
order-left:1pt solid rgb(204,204,204);padding:0in 0in 0in 6pt;margin:5pt 0i=
n 5pt 4.8pt">
<div>
<p class=3D"MsoNormal">I was agreeing with MT, but I&#39;m happy to=C2=A0se=
e some more MUSTs added if people feel that&#39;d be helpful.<u></u><u></u>=
</p>
</div>
</blockquote>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt">=C2=A0</span><u></u><=
u></u></p>
<div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt">Coincidentally, we we=
re just talking about this internally at Akamai yesterday. IMO, an invarian=
ts document isn&#39;t really helpful if it isn&#39;t normative,
 and for it to be normative it (or a related practices doc for operators) r=
eally needs to spell out clear boundaries for operators with MUSTs..</span>=
<u></u><u></u></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt">=C2=A0</span><u></u><=
u></u></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt">The example that came=
 up yesterday was around operators filtering QUIC in the event of a DDoS: o=
ne recommendation based on some conversations going
 back at least to Prague 2019 was to hash packets on 4-tuple and filter tho=
se below a hash value chosen for a desired ingress limit instead of doing w=
hat most operators do with UDP today, which is to cap UDP throughput and ju=
st drop packets randomly or tail
 drop.</span><u></u><u></u></p>
</div>
</div>
</div>
</div>
</blockquote>
<p>Interesting. Did they consider using the CID, or a fraction of it? This =
looks entirely like the scenario for which we developed stateless retry.<u>=
</u><u></u></p>
<blockquote style=3D"margin-top:5pt;margin-bottom:5pt">
<div>
<div>
<div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt">=C2=A0</span><u></u><=
u></u></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt">This recommendation c=
ertainly imposes some constraints on future protocol development that motiv=
ate new invariants: for instance, it would preclude
 sharding a connection across multiple source ports (not that there is nece=
ssarily a reason to do this; it&#39;s just an example). But more importantl=
y, it goes beyond invariants: it&#39;s one among many practices compatible =
with the current set of invariants, some
 reasonable and some terrible.</span><u></u><u></u></p>
</div>
</div>
</div>
</div>
</blockquote>
<p>This would break the &quot;preferred address&quot; redirection. Preferre=
d address migration may or may not be spelled out in the invariants.<u></u>=
<u></u></p>
<blockquote style=3D"margin-top:5pt;margin-bottom:5pt">
<div>
<div>
<div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt">=C2=A0</span><u></u><=
u></u></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:12pt">Operators are going t=
o do things to QUIC traffic, so it would be good to offer them recommendati=
ons that are compatible with broad deployability.</span><u></u><u></u></p>
</div>
</div>
</div>
</div>
</blockquote>
<p>=C2=A0<u></u><u></u></p>
<p>Yes, we do need the invariants for that.<u></u><u></u></p>
<p>-- Christian Huitema<u></u><u></u></p>
</div>
</div>
</div>
</blockquote>
</div>
</div>
</div>
</div>

</blockquote></div>

--000000000000f29bcc05a874461c--


From nobody Fri Jun 19 13:42:29 2020
Return-Path: <ilubashe@akamai.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 7B67F3A0E71 for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 13:42:28 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.999
X-Spam-Level: 
X-Spam-Status: No, score=-1.999 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, HTML_MESSAGE=0.001, HTTPS_HTTP_MISMATCH=0.1, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=unavailable autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=akamai.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ZnGusEDUTeBs for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 13:42:26 -0700 (PDT)
Received: from mx0a-00190b01.pphosted.com (mx0a-00190b01.pphosted.com [IPv6:2620:100:9001:583::1]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 394203A0E6D for <quic@ietf.org>; Fri, 19 Jun 2020 13:42:26 -0700 (PDT)
Received: from pps.filterd (m0122332.ppops.net [127.0.0.1]) by mx0a-00190b01.pphosted.com (8.16.0.42/8.16.0.42) with SMTP id 05JKcLmv002348; Fri, 19 Jun 2020 21:42:16 +0100
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=akamai.com; h=from : to : cc : subject : date : message-id : references : in-reply-to : content-type : mime-version; s=jan2016.eng; bh=OjpyhO0WCNaWTdPH4Sv/gn1wlqmljjyv4gcTIN+39dA=; b=II3Q2yPr8SEqsSpAzwqCopGYtHjo+ozdS1jS6tT9BBTfx+Idko1lgFsXAoEcuON1E6sf fqkLHV5etjVgFkBDG52XBfea+6NNzcCLU++LJprFMMk+IQ2cTRULpH3TMdVcEhrj4ZZe n38X5RKcPqwJ7oyoFoqaJsXj8PBWgo+ctyoQwYCFM48py0fb6Zj3Q6/c/cXThxTlfG5v YPKuu37jmcC7lJrSj+jX9Ou0dRddsyy1UYQ9CtbhMCnp2nrXPoz/RMwqS72CPHo/UgF/ EY2PYMr7QTFH5aNcKkFsT1An9zUh+M52tMmg9IK2N8vnc4E1Tj3CXOpIqfZbd0GZhTUO EA== 
Received: from prod-mail-ppoint1 (prod-mail-ppoint1.akamai.com [184.51.33.18] (may be forged)) by mx0a-00190b01.pphosted.com with ESMTP id 31qqyc5gy9-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Fri, 19 Jun 2020 21:42:15 +0100
Received: from pps.filterd (prod-mail-ppoint1.akamai.com [127.0.0.1]) by prod-mail-ppoint1.akamai.com (8.16.0.42/8.16.0.42) with SMTP id 05JKZc33022889; Fri, 19 Jun 2020 16:42:14 -0400
Received: from email.msg.corp.akamai.com ([172.27.123.31]) by prod-mail-ppoint1.akamai.com with ESMTP id 31qjmbwt33-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-SHA384 bits=256 verify=NOT); Fri, 19 Jun 2020 16:42:14 -0400
Received: from USMA1EX-DAG1MB5.msg.corp.akamai.com (172.27.123.105) by usma1ex-dag1mb4.msg.corp.akamai.com (172.27.123.104) with Microsoft SMTP Server (TLS) id 15.0.1497.2; Fri, 19 Jun 2020 16:42:13 -0400
Received: from USMA1EX-DAG1MB5.msg.corp.akamai.com ([172.27.123.105]) by usma1ex-dag1mb5.msg.corp.akamai.com ([172.27.123.105]) with mapi id 15.00.1497.006; Fri, 19 Jun 2020 16:42:13 -0400
From: "Lubashev, Igor" <ilubashe@akamai.com>
To: Martin Duke <martin.h.duke@gmail.com>
CC: Christian Huitema <huitema@huitema.net>, Kyle Rose <krose@krose.org>, "Ian Swett" <ianswett=40google.com@dmarc.ietf.org>, Lars Eggert <lars@eggert.org>, IETF QUIC WG <quic@ietf.org>, Jared Mauch <jared@puck.nether.net>
Subject: RE: Is the invariants draft really standards track?
Thread-Topic: Is the invariants draft really standards track?
Thread-Index: AQHWM4bg7m/IhU9tKEaMc9qs7omEcqi7wjeAgACBhgCAAA8vAIAAAZ8A///Fg1CAId32AIACNYSAgACBoAD//9wAkA==
Date: Fri, 19 Jun 2020 20:42:13 +0000
Message-ID: <9c2e300c30f74d1794d11cf4334ea07b@usma1ex-dag1mb5.msg.corp.akamai.com>
References: <CAM4esxQBqfrz24riPQA_VGKcGp_TzW0pqb97KfFMtNdW9pUfDg@mail.gmail.com> <833A693C-62E6-4889-9954-FCE65A839A7C@eggert.org> <CAKcm_gPMO2DtqvKucqVw0zDjSniSOmFD4p1Tp4YLjr9WSWdEUw@mail.gmail.com> <CAJU8_nUN42gGmQof24XD9-EjXedyzcarDyRP8MGe1qW-BZ=+Aw@mail.gmail.com> <9cd91c24-c730-22a4-7aa0-baf61613b3ce@huitema.net> <f4922cdb59014202900de44cc5fea0ff@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQvwkTvpUcu6-+W5zWo22m-R1jvN7DcCpXfuw8Hb55qsw@mail.gmail.com> <95dd02c92b32472d9cab0dd47b98c637@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQxxXn27rZEY75-AsHD5VF0fqiV1VDyeSrzQ=-sM7JNCg@mail.gmail.com>
In-Reply-To: <CAM4esxQxxXn27rZEY75-AsHD5VF0fqiV1VDyeSrzQ=-sM7JNCg@mail.gmail.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
x-ms-exchange-transport-fromentityheader: Hosted
x-originating-ip: [172.19.118.2]
Content-Type: multipart/alternative; boundary="_000_9c2e300c30f74d1794d11cf4334ea07busma1exdag1mb5msgcorpak_"
MIME-Version: 1.0
X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.216, 18.0.687 definitions=2020-06-19_22:2020-06-19, 2020-06-19 signatures=0
X-Proofpoint-Spam-Details: rule=notspam policy=default score=0 spamscore=0 malwarescore=0 bulkscore=0 adultscore=0 mlxlogscore=999 mlxscore=0 phishscore=0 suspectscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2004280000 definitions=main-2006190146
X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.216, 18.0.687 definitions=2020-06-19_22:2020-06-19, 2020-06-19 signatures=0
X-Proofpoint-Spam-Details: rule=notspam policy=default score=0 spamscore=0 malwarescore=0 bulkscore=0 lowpriorityscore=0 impostorscore=0 priorityscore=1501 phishscore=0 suspectscore=0 mlxlogscore=999 mlxscore=0 clxscore=1015 adultscore=0 cotscore=-2147483648 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2004280000 definitions=main-2006190146
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/mUHxp00uj15kBoImrVVUqLb2Rkg>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 19 Jun 2020 20:42:29 -0000

--_000_9c2e300c30f74d1794d11cf4334ea07busma1exdag1mb5msgcorpak_
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_9c2e300c30f74d1794d11cf4334ea07busma1exdag1mb5msgcorpak_
Content-Type: text/html; charset="utf-8"
Content-Transfer-Encoding: base64
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--_000_9c2e300c30f74d1794d11cf4334ea07busma1exdag1mb5msgcorpak_--


From nobody Fri Jun 19 14:49:33 2020
Return-Path: <huitema@huitema.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id E33663A0EE8 for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 14:49:24 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.798
X-Spam-Level: 
X-Spam-Status: No, score=-1.798 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, HTTPS_HTTP_MISMATCH=0.1, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=no autolearn_force=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id g8WThokGGm9Q for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 14:49:22 -0700 (PDT)
Received: from mx36-out10.antispamcloud.com (mx36-out10.antispamcloud.com [209.126.121.30]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id CA19E3A0EB6 for <quic@ietf.org>; Fri, 19 Jun 2020 14:49:21 -0700 (PDT)
Received: from xse405.mail2web.com ([66.113.197.151] helo=xse.mail2web.com) by mx18.antispamcloud.com with esmtp (Exim 4.92) (envelope-from <huitema@huitema.net>) id 1jmOsr-0008IA-8N for quic@ietf.org; Fri, 19 Jun 2020 23:49:20 +0200
Received: from xsmtp22.mail2web.com (unknown [10.100.68.61]) by xse.mail2web.com (Postfix) with ESMTPS id 49pXW51JSPz8XK6 for <quic@ietf.org>; Fri, 19 Jun 2020 14:48:49 -0700 (PDT)
Received: from [10.5.2.17] (helo=xmail07.myhosting.com) by xsmtp22.mail2web.com with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:256) (Exim 4.92) (envelope-from <huitema@huitema.net>) id 1jmOsT-0000Jd-10 for quic@ietf.org; Fri, 19 Jun 2020 14:48:49 -0700
Received: (qmail 19804 invoked from network); 19 Jun 2020 21:48:48 -0000
Received: from unknown (HELO [192.168.1.107]) (Authenticated-user:_huitema@huitema.net@[172.58.43.64]) (envelope-sender <huitema@huitema.net>) by xmail07.myhosting.com (qmail-ldap-1.03) with ESMTPA for <quic@ietf.org>; 19 Jun 2020 21:48:48 -0000
To: "Lubashev, Igor" <ilubashe=40akamai.com@dmarc.ietf.org>, Martin Duke <martin.h.duke@gmail.com>
Cc: Jared Mauch <jared@puck.nether.net>, Kyle Rose <krose@krose.org>, Ian Swett <ianswett=40google.com@dmarc.ietf.org>, Lars Eggert <lars@eggert.org>, IETF QUIC WG <quic@ietf.org>
References: <CAM4esxQBqfrz24riPQA_VGKcGp_TzW0pqb97KfFMtNdW9pUfDg@mail.gmail.com> <833A693C-62E6-4889-9954-FCE65A839A7C@eggert.org> <CAKcm_gPMO2DtqvKucqVw0zDjSniSOmFD4p1Tp4YLjr9WSWdEUw@mail.gmail.com> <CAJU8_nUN42gGmQof24XD9-EjXedyzcarDyRP8MGe1qW-BZ=+Aw@mail.gmail.com> <9cd91c24-c730-22a4-7aa0-baf61613b3ce@huitema.net> <f4922cdb59014202900de44cc5fea0ff@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQvwkTvpUcu6-+W5zWo22m-R1jvN7DcCpXfuw8Hb55qsw@mail.gmail.com> <95dd02c92b32472d9cab0dd47b98c637@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQxxXn27rZEY75-AsHD5VF0fqiV1VDyeSrzQ=-sM7JNCg@mail.gmail.com> <9c2e300c30f74d1794d11cf4334ea07b@usma1ex-dag1mb5.msg.corp.akamai.com>
From: Christian Huitema <huitema@huitema.net>
Autocrypt: addr=huitema@huitema.net; prefer-encrypt=mutual; keydata= mDMEXtavGxYJKwYBBAHaRw8BAQdA1ou9A5MHTP9N3jfsWzlDZ+jPnQkusmc7sfLmWVz1Rmu0 J0NocmlzdGlhbiBIdWl0ZW1hIDxodWl0ZW1hQGh1aXRlbWEubmV0PoiWBBMWCAA+FiEEw3G4 Nwi4QEpAAXUUELAmqKBYtJQFAl7WrxsCGwMFCQlmAYAFCwkIBwIGFQoJCAsCBBYCAwECHgEC F4AACgkQELAmqKBYtJQbMwD/ebj/qnSbthC/5kD5DxZ/Ip0CGJw5QBz/+fJp3R8iAlsBAMjK r2tmyWyJz0CUkVG24WaR5EAJDvgwDv8h22U6QVkAuDgEXtavGxIKKwYBBAGXVQEFAQEHQJoM 6MUAIqpoqdCIiACiEynZf7nlJg2Eu0pXIhbUGONdAwEIB4h+BBgWCAAmFiEEw3G4Nwi4QEpA AXUUELAmqKBYtJQFAl7WrxsCGwwFCQlmAYAACgkQELAmqKBYtJRm2wD7BzeK5gEXSmBcBf0j BYdSaJcXNzx4yPLbP4GnUMAyl2cBAJzcsR4RkwO4dCRqM9CHpVJCwHtbUDJaa55//E0kp+gH
Subject: Re: Is the invariants draft really standards track?
Message-ID: <2c40f3d9-fa40-9834-ac30-36bc9a1a6303@huitema.net>
Date: Fri, 19 Jun 2020 14:48:48 -0700
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:68.0) Gecko/20100101 Thunderbird/68.9.0
MIME-Version: 1.0
In-Reply-To: <9c2e300c30f74d1794d11cf4334ea07b@usma1ex-dag1mb5.msg.corp.akamai.com>
Content-Type: multipart/alternative; boundary="------------78BEF07D38E0F11540D1141A"
Content-Language: en-US
X-Originating-IP: 66.113.197.151
X-Spampanel-Domain: xsmtpout.mail2web.com
X-Spampanel-Username: 66.113.197.0/24
Authentication-Results: antispamcloud.com; auth=pass smtp.auth=66.113.197.0/24@xsmtpout.mail2web.com
X-Spampanel-Outgoing-Class: unsure
X-Spampanel-Outgoing-Evidence: Combined (0.15)
X-Recommended-Action: accept
X-Filter-ID: Mvzo4OR0dZXEDF/gcnlw0f6LF1GdvkEexklpcFpSF5apSDasLI4SayDByyq9LIhVUZbR67CQ7/vm /hHDJU4RXkTNWdUk1Ol2OGx3IfrIJKywOmJyM1qr8uRnWBrbSAGDcnqpk5VeF3xR4kF6iVwRtbgN zB/4Jkrw1eDLcif59ftWVWQDBUNR1RlC0Cuws2FOU7Tmz6iKnkQL9gqsxD347235Nhqq+/HvroPq 8GSPg+5hmwN8D4LrepG7AX8WNwY8PM2DmctUkuqITdPZ8a3LuY6jSvfpO+1kZkomjtjB6X5Q5Q9f RUeIpTIC2ySfqvnqLwoxlgatmaBb0rBiK9xbkDrUqzcKIief90MVLZY9LbIZh9+IQ1oS9LBn3VIP 95Jz7ujRlJ9wSMlhvaudJXZ9EIBG/qaR+8r9SKFMmPJLf850OvZYsmoVQuOIhwKLK6IKBNB4LZ0v UHHKTzJX7b1JhLSQQ4vSj0QEim26t/Moy0UPX5E73H1QfrH/5kkrV/Cr0bm2vWdo8usP65i82q1C dZgGrpL44wdx9eXqjQjbvUopOMQJvQ/Ck3iiU+4DQAj3fuQgzT3K9JUHTNiGwfwAmxx/Wk8McinP JEkgAVrOMpbIPHELHGuybVREcKQKsZ6S4LcBfuQ2UTwAgewOf3yUBSxCkcZUXdAZaPbURRvf5Ecv UwPy3x0FYtCNEb10sHyQCLHEvD1OqP6bgZ4L66GcgBg66gs5OuzYxJgw5atIxeNDvjI/CYe5WPy0 +t1RP0azMDzivn+U4bsKYbUeciCcn5xMPnetLBJMh51NiRRoHIAWe52rGzfY483zGYnBiem0miK7 x42VjdzChZMe6O/Did+/hGXTmfhE+Dx2/NyzMXrFCkQ40uY20UVTgqg+hJ1IvOnCUbNPgcPcQwzM gKHyQxUo+ql2ySTkvEFH/23XMww2BnTTFGX5/yI4Ky+1ZJcbGqc5H4PEZHeoI/d6LWFf332z7LMw LGdoi9FMQ5j9dQUvMi1YKAun15JQSJLyCT5k+MTObVKxHy/dols381l9r9ft9daDonlwd6LnuX+J u10=
X-Report-Abuse-To: spam@quarantine11.antispamcloud.com
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/G7PI8Mr5DjCyyHSibYIUSMYpE8U>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 19 Jun 2020 21:49:31 -0000

This is a multi-part message in MIME format.
--------------78BEF07D38E0F11540D1141A
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable

When under DOS attack, you want to "minimize blowback", i.e., as much as
possible avoid generating packets in response to attack traffic. So,
yes, a server may choose to not send stateless resets to anyone when
under attack; in fact, my recommendation would be that a server SHOULD
NOT send stateless resets to anyone when under attack.

That said, Igor raised an interesting point about return traffic. It
would be very nice if DOS protection boxes could distinguish between
"validated traffic" that the server presumably intends to process, and
"unsolicited traffic" that will just consume resource. The box can then
reserve some share of the resource for validated traffic, and place the
rest of the traffic in a lower priority queue. Fine, but there needs to
be a test. The classic test is that incoming traffic is "validated" if
the protection box can match it with return traffic coming from the
server -- for some definition of matching.

=46rom that point of view, stateless reset is definitely not helpful. But=

problematic traffic goes beyond that. The server will reply to a
client's initial with a server's initial packet. Does that validate the
response traffic? OK, maybe the protection box can programmed to only
validate traffic if it sees 1RTT packets. But many servers will send
1-RTT packets as 0.5 RTT. Does that validate the response traffic?

We might say that traffic is validated when the handshake is confirmed,
but the protection box does not understand the TLS handshake, it just
sees packet types and packet sizes. It cannot distinguish between 0.5RTT
data and 1RTT data, and thus the closest approximation of "validation"
would be seeing more than an initial window worth of traffic coming back
from the server. That does not sound great.

On the other hand, things get much better if the server under attack can
adopt a defensive posture and help the DOS protection box do its job.
Suppose that a server can detect that it is under attack -- or be
explicitly configured so. The simplest defensive posture would be to (1)
disable stateless reset and (2) not send any 0.5RTT packet, including in
response to 0-RTT. The protection boxes can at that point take the 1-RTT
packets from the server as indicating validation.

Maybe we should specify that.

-- Christian Huitema

On 6/19/2020 1:42 PM, Lubashev, Igor wrote:
>
> > There is no need for servers to decrypt CIDs in QUIC-LB. Presumably
> the server has a lookup table for its CIDs.
>
> =C2=A0
>
> Sending a stateless reset in response to a junk packet would cost more
> CPU than verifying CID integrity.=C2=A0 But, yes, a server may choose t=
o
> not send stateless resets to anyone when under attack.
>
> =C2=A0
>
> =C2=A0
>
> *From:* Martin Duke <martin.h.duke@gmail.com>
> *Sent:* Friday, June 19, 2020 2:44 PM
>
> >Unfortunately, Retry system protects only server's memory state and
> some CPU cycles spent on crypto.=C2=A0 (Servers still need to decrypt C=
ID
> to decide it is invalid, and if the attacker is clever enough to
> establish one valid connection and use that CID in a flood, the server
> will also be decrypting packets.)=C2=A0=C2=A0
>
> =C2=A0
>
> There is no need for servers to decrypt CIDs in QUIC-LB. Presumably
> the server has a lookup table for its CIDs.
>
> =C2=A0
>
> It is true that Retry Services (and indeed, the Retry concept as a
> whole) does nothing to protect network capacity.
>
> =C2=A0
>
> On Fri, Jun 19, 2020 at 8:08 AM Lubashev, Igor <ilubashe@akamai.com
> <mailto:ilubashe@akamai.com>> wrote:
>
>     It looks like
>     https://tools.ietf.org/html/draft-ietf-quic-load-balancers-02#secti=
on-5
>     <https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__tools.ietf.=
org_html_draft-2Dietf-2Dquic-2Dload-2Dbalancers-2D02-23section-2D5&d=3DDw=
MFaQ&c=3D96ZbZZcaMF4w0F4jpN6LZg&r=3DDjn3bQ5uNJDPM_2skfL3rW1tzcIxyjUZdn_m5=
5KPmlo&m=3DeTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTKljvj-ys&s=3D1x7AN2a51X_zV5=
xSyK0uCL8vZ5cugD3n4lbFWZDqVQg&e=3D>
>     is an excellent discussion of Retry mechanics.=C2=A0 It definitely
>     deserves a reference from this manageability draft.
>
>     =C2=A0
>
>     The Retry mechanisms described in LB draft are all cooperating
>     boxes, and servers must be aware of them.=C2=A0 Unfortunately, Retr=
y
>     system protects only server's memory state and some CPU cycles
>     spent on crypto.=C2=A0 (Servers still need to decrypt CID to decide=
 it
>     is invalid, and if the attacker is clever enough to establish one
>     valid connection and use that CID in a flood, the server will also
>     be decrypting packets.)=C2=A0 Retry does nothing to protect network=

>     resources.
>
>     =C2=A0
>
>     The PR I opened (https://github.com/quicwg/ops-drafts/pull/94
>     <https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__github.com_=
quicwg_ops-2Ddrafts_pull_94&d=3DDwMFaQ&c=3D96ZbZZcaMF4w0F4jpN6LZg&r=3DDjn=
3bQ5uNJDPM_2skfL3rW1tzcIxyjUZdn_m55KPmlo&m=3DeTT-BoZ1fMitywKRVSxpU3js0lhO=
0qkspYTKljvj-ys&s=3DVpvhQ9n79rANIk3O5Sm7PZyToFQEennoPt9iqFpWbq8&e=3D>)
>     is about uncooperating devices that try to mitigate volumetric
>     network attacks.
>
>     =C2=A0
>
>     =C2=A0
>
>     *From:* Martin Duke <martin.h.duke@gmail.com
>     <mailto:martin.h.duke@gmail.com>>
>     *Sent:* Wednesday, June 17, 2020 8:16 PM
>
>     Hi Igor, you might want to check out the "Retry Services" bit of
>     the QUIC-LB draft. This has something to do with the DDoS use case
>     you discuss.
>
>     =C2=A0
>
>     On Wed, May 27, 2020 at 9:07 AM Lubashev, Igor
>     <ilubashe@akamai.com <mailto:ilubashe@akamai.com>> wrote:
>
>         I=E2=80=99m working on a manageability draft PR for this (how t=
o rate
>         limit UDP to reduce disruption to QUIC if you have to rate
>         limit UDP).=C2=A0 ETA end of the week (if I do not get pulled i=
nto
>         something again).
>
>         =C2=A0
>
>         The relevant observation is that DDoS with UDP that is
>         indistinguishable from QUIC will happen.=C2=A0 UDP is already t=
he
>         most prevalent DDoS vector, since it is easy for a compromised
>         non-admin app to send a flood of huge UDP packets (with TCP
>         you get throttled by the congestion controller).=C2=A0 So there=

>         WILL be DDoS protection devices out there to try to mitigate
>         the problem, possibly by observing both directions of the flow
>         and deciding whether a packet belongs to a valid flow or not.
>
>         =C2=A0
>
>         Since such middle boxes will be created, the more explicit and
>         normative Invariants are about what one can expect, the less
>         such middle boxes may decide for themselves.=C2=A0 For example =
(I
>         did not think long about it), if some elements of path
>         validation could land into Invariants (roughly, =E2=80=9Cno mor=
e than
>         X packets/bytes can be sent on a new path w/o a return
>         packet=E2=80=9D), a DDoS middle box may use this fact and activ=
e
>         connection migration might still have a chance during an
>         attack (NAT rebinding could be linked by DDoS boxes to an old
>         connection via unchanged CID).
>
>         =C2=A0
>
>           * Igor
>
>         =C2=A0
>
>         =C2=A0
>
>         *From:* Christian Huitema <huitema@huitema.net
>         <mailto:huitema@huitema.net>>
>         *Sent:* Wednesday, May 27, 2020 11:34 AM
>
>         On 5/27/2020 8:28 AM, Kyle Rose wrote:
>
>             On Wed, May 27, 2020 at 10:34 AM Ian Swett
>             <ianswett=3D40google.com@dmarc.ietf.org
>             <mailto:40google.com@dmarc.ietf.org>> wrote:
>
>                 I was agreeing with MT, but I'm happy to=C2=A0see some =
more
>                 MUSTs added if people feel that'd be helpful.
>
>             =C2=A0
>
>             Coincidentally, we were just talking about this internally
>             at Akamai yesterday. IMO, an invariants document isn't
>             really helpful if it isn't normative, and for it to be
>             normative it (or a related practices doc for operators)
>             really needs to spell out clear boundaries for operators
>             with MUSTs..
>
>             =C2=A0
>
>             The example that came up yesterday was around operators
>             filtering QUIC in the event of a DDoS: one recommendation
>             based on some conversations going back at least to Prague
>             2019 was to hash packets on 4-tuple and filter those below
>             a hash value chosen for a desired ingress limit instead of
>             doing what most operators do with UDP today, which is to
>             cap UDP throughput and just drop packets randomly or tail
>             drop.
>
>         Interesting. Did they consider using the CID, or a fraction of
>         it? This looks entirely like the scenario for which we
>         developed stateless retry.
>
>             =C2=A0
>
>             This recommendation certainly imposes some constraints on
>             future protocol development that motivate new invariants:
>             for instance, it would preclude sharding a connection
>             across multiple source ports (not that there is
>             necessarily a reason to do this; it's just an example).
>             But more importantly, it goes beyond invariants: it's one
>             among many practices compatible with the current set of
>             invariants, some reasonable and some terrible.
>
>         This would break the "preferred address" redirection.
>         Preferred address migration may or may not be spelled out in
>         the invariants.
>
>             =C2=A0
>
>             Operators are going to do things to QUIC traffic, so it
>             would be good to offer them recommendations that are
>             compatible with broad deployability.
>
>         =C2=A0
>
>         Yes, we do need the invariants for that.
>
>         -- Christian Huitema
>

--------------78BEF07D38E0F11540D1141A
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: 8bit

<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <p>When under DOS attack, you want to "minimize blowback", i.e., as
      much as possible avoid generating packets in response to attack
      traffic. So, yes, a server may choose to not send stateless resets
      to anyone when under attack; in fact, my recommendation would be
      that a server SHOULD NOT send stateless resets to anyone when
      under attack.</p>
    <p>That said, Igor raised an interesting point about return traffic.
      It would be very nice if DOS protection boxes could distinguish
      between "validated traffic" that the server presumably intends to
      process, and "unsolicited traffic" that will just consume
      resource. The box can then reserve some share of the resource for
      validated traffic, and place the rest of the traffic in a lower
      priority queue. Fine, but there needs to be a test. The classic
      test is that incoming traffic is "validated" if the protection box
      can match it with return traffic coming from the server -- for
      some definition of matching.</p>
    <p>From that point of view, stateless reset is definitely not
      helpful. But problematic traffic goes beyond that. The server will
      reply to a client's initial with a server's initial packet. Does
      that validate the response traffic? OK, maybe the protection box
      can programmed to only validate traffic if it sees 1RTT packets.
      But many servers will send 1-RTT packets as 0.5 RTT. Does that
      validate the response traffic?</p>
    <p>We might say that traffic is validated when the handshake is
      confirmed, but the protection box does not understand the TLS
      handshake, it just sees packet types and packet sizes. It cannot
      distinguish between 0.5RTT data and 1RTT data, and thus the
      closest approximation of "validation" would be seeing more than an
      initial window worth of traffic coming back from the server. That
      does not sound great.</p>
    <p>On the other hand, things get much better if the server under
      attack can adopt a defensive posture and help the DOS protection
      box do its job. Suppose that a server can detect that it is under
      attack -- or be explicitly configured so. The simplest defensive
      posture would be to (1) disable stateless reset and (2) not send
      any 0.5RTT packet, including in response to 0-RTT. The protection
      boxes can at that point take the 1-RTT packets from the server as
      indicating validation.</p>
    <p>Maybe we should specify that.</p>
    <p>-- Christian Huitema<br>
    </p>
    <div class="moz-cite-prefix">On 6/19/2020 1:42 PM, Lubashev, Igor
      wrote:<br>
    </div>
    <blockquote type="cite"
cite="mid:9c2e300c30f74d1794d11cf4334ea07b@usma1ex-dag1mb5.msg.corp.akamai.com">
      <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
      <meta name="Generator" content="Microsoft Word 15 (filtered
        medium)">
      <style><!--
/* Font Definitions */
@font-face
	{font-family:"Cambria Math";
	panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
	{font-family:Calibri;
	panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
	{margin:0in;
	margin-bottom:.0001pt;
	font-size:11.0pt;
	font-family:"Calibri",sans-serif;}
a:link, span.MsoHyperlink
	{mso-style-priority:99;
	color:blue;
	text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
	{mso-style-priority:99;
	color:purple;
	text-decoration:underline;}
p.msonormal0, li.msonormal0, div.msonormal0
	{mso-style-name:msonormal;
	mso-margin-top-alt:auto;
	margin-right:0in;
	mso-margin-bottom-alt:auto;
	margin-left:0in;
	font-size:11.0pt;
	font-family:"Calibri",sans-serif;}
span.EmailStyle19
	{mso-style-type:personal-reply;
	font-family:"Calibri",sans-serif;
	color:windowtext;}
.MsoChpDefault
	{mso-style-type:export-only;
	font-family:"Calibri",sans-serif;}
@page WordSection1
	{size:8.5in 11.0in;
	margin:1.0in 1.0in 1.0in 1.0in;}
div.WordSection1
	{page:WordSection1;}
/* List Definitions */
@list l0
	{mso-list-id:554779372;
	mso-list-template-ids:587751912;}
@list l0:level1
	{mso-level-number-format:bullet;
	mso-level-text:;
	mso-level-tab-stop:.5in;
	mso-level-number-position:left;
	text-indent:-.25in;
	mso-ansi-font-size:10.0pt;
	font-family:Symbol;}
@list l0:level2
	{mso-level-number-format:bullet;
	mso-level-text:;
	mso-level-tab-stop:1.0in;
	mso-level-number-position:left;
	text-indent:-.25in;
	mso-ansi-font-size:10.0pt;
	font-family:Symbol;}
@list l0:level3
	{mso-level-number-format:bullet;
	mso-level-text:;
	mso-level-tab-stop:1.5in;
	mso-level-number-position:left;
	text-indent:-.25in;
	mso-ansi-font-size:10.0pt;
	font-family:Symbol;}
@list l0:level4
	{mso-level-number-format:bullet;
	mso-level-text:;
	mso-level-tab-stop:2.0in;
	mso-level-number-position:left;
	text-indent:-.25in;
	mso-ansi-font-size:10.0pt;
	font-family:Symbol;}
@list l0:level5
	{mso-level-number-format:bullet;
	mso-level-text:;
	mso-level-tab-stop:2.5in;
	mso-level-number-position:left;
	text-indent:-.25in;
	mso-ansi-font-size:10.0pt;
	font-family:Symbol;}
@list l0:level6
	{mso-level-number-format:bullet;
	mso-level-text:;
	mso-level-tab-stop:3.0in;
	mso-level-number-position:left;
	text-indent:-.25in;
	mso-ansi-font-size:10.0pt;
	font-family:Symbol;}
@list l0:level7
	{mso-level-number-format:bullet;
	mso-level-text:;
	mso-level-tab-stop:3.5in;
	mso-level-number-position:left;
	text-indent:-.25in;
	mso-ansi-font-size:10.0pt;
	font-family:Symbol;}
@list l0:level8
	{mso-level-number-format:bullet;
	mso-level-text:;
	mso-level-tab-stop:4.0in;
	mso-level-number-position:left;
	text-indent:-.25in;
	mso-ansi-font-size:10.0pt;
	font-family:Symbol;}
@list l0:level9
	{mso-level-number-format:bullet;
	mso-level-text:;
	mso-level-tab-stop:4.5in;
	mso-level-number-position:left;
	text-indent:-.25in;
	mso-ansi-font-size:10.0pt;
	font-family:Symbol;}
ol
	{margin-bottom:0in;}
ul
	{margin-bottom:0in;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext="edit" spidmax="1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext="edit">
<o:idmap v:ext="edit" data="1" />
</o:shapelayout></xml><![endif]-->
      <div class="WordSection1">
        <p class="MsoNormal">&gt; There is no need for servers to
          decrypt CIDs in QUIC-LB. Presumably the server has a lookup
          table for its CIDs.<o:p></o:p></p>
        <p class="MsoNormal"><o:p> </o:p></p>
        <p class="MsoNormal">Sending a stateless reset in response to a
          junk packet would cost more CPU than verifying CID integrity. 
          But, yes, a server may choose to not send stateless resets to
          anyone when under attack.<o:p></o:p></p>
        <p class="MsoNormal"><o:p> </o:p></p>
        <p class="MsoNormal"><o:p> </o:p></p>
        <div style="border:none;border-left:solid blue 1.5pt;padding:0in
          0in 0in 4.0pt">
          <div>
            <div style="border:none;border-top:solid #E1E1E1
              1.0pt;padding:3.0pt 0in 0in 0in">
              <p class="MsoNormal"><b>From:</b> Martin Duke
                <a class="moz-txt-link-rfc2396E" href="mailto:martin.h.duke@gmail.com">&lt;martin.h.duke@gmail.com&gt;</a> <br>
                <b>Sent:</b> Friday, June 19, 2020 2:44 PM<br>
                <br>
                <o:p></o:p></p>
            </div>
          </div>
          <div>
            <p class="MsoNormal">&gt;Unfortunately, Retry system
              protects only server's memory state and some CPU cycles
              spent on crypto.  (Servers still need to decrypt CID to
              decide it is invalid, and if the attacker is clever enough
              to establish one valid connection and use that CID in a
              flood, the server will also be decrypting packets.)  <o:p></o:p></p>
            <div>
              <p class="MsoNormal"><o:p> </o:p></p>
            </div>
            <div>
              <p class="MsoNormal">There is no need for servers to
                decrypt CIDs in QUIC-LB. Presumably the server has a
                lookup table for its CIDs.<o:p></o:p></p>
            </div>
            <div>
              <p class="MsoNormal"><o:p> </o:p></p>
            </div>
            <div>
              <p class="MsoNormal">It is true that Retry Services (and
                indeed, the Retry concept as a whole) does nothing to
                protect network capacity.<o:p></o:p></p>
            </div>
          </div>
          <p class="MsoNormal"><o:p> </o:p></p>
          <div>
            <div>
              <p class="MsoNormal">On Fri, Jun 19, 2020 at 8:08 AM
                Lubashev, Igor &lt;<a href="mailto:ilubashe@akamai.com"
                  moz-do-not-send="true">ilubashe@akamai.com</a>&gt;
                wrote:<o:p></o:p></p>
            </div>
            <blockquote style="border:none;border-left:solid #CCCCCC
              1.0pt;padding:0in 0in 0in
              6.0pt;margin-left:4.8pt;margin-right:0in">
              <div>
                <div>
                  <p class="MsoNormal"
                    style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto">It
                    looks like
                    <a
href="https://urldefense.proofpoint.com/v2/url?u=https-3A__tools.ietf.org_html_draft-2Dietf-2Dquic-2Dload-2Dbalancers-2D02-23section-2D5&amp;d=DwMFaQ&amp;c=96ZbZZcaMF4w0F4jpN6LZg&amp;r=Djn3bQ5uNJDPM_2skfL3rW1tzcIxyjUZdn_m55KPmlo&amp;m=eTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTKljvj-ys&amp;s=1x7AN2a51X_zV5xSyK0uCL8vZ5cugD3n4lbFWZDqVQg&amp;e="
                      target="_blank" moz-do-not-send="true">
https://tools.ietf.org/html/draft-ietf-quic-load-balancers-02#section-5</a>
                    is an excellent discussion of Retry mechanics.  It
                    definitely deserves a reference from this
                    manageability draft.<o:p></o:p></p>
                  <p class="MsoNormal"
                    style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"> <o:p></o:p></p>
                  <p class="MsoNormal"
                    style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto">The
                    Retry mechanisms described in LB draft are all
                    cooperating boxes, and servers must be aware of
                    them.  Unfortunately, Retry system protects only
                    server's memory state and some CPU cycles spent on
                    crypto.  (Servers still need to decrypt CID to
                    decide it is invalid, and if the attacker is clever
                    enough to establish one valid connection and use
                    that CID in a flood, the server will also be
                    decrypting packets.)  Retry does nothing to protect
                    network resources.<o:p></o:p></p>
                  <p class="MsoNormal"
                    style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"> <o:p></o:p></p>
                  <p class="MsoNormal"
                    style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto">The
                    PR I opened (<a
href="https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_quicwg_ops-2Ddrafts_pull_94&amp;d=DwMFaQ&amp;c=96ZbZZcaMF4w0F4jpN6LZg&amp;r=Djn3bQ5uNJDPM_2skfL3rW1tzcIxyjUZdn_m55KPmlo&amp;m=eTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTKljvj-ys&amp;s=VpvhQ9n79rANIk3O5Sm7PZyToFQEennoPt9iqFpWbq8&amp;e="
                      target="_blank" moz-do-not-send="true">https://github.com/quicwg/ops-drafts/pull/94</a>)
                    is about uncooperating devices that try to mitigate
                    volumetric network attacks.<o:p></o:p></p>
                  <p class="MsoNormal"
                    style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"> <o:p></o:p></p>
                  <p class="MsoNormal"
                    style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"> <o:p></o:p></p>
                  <div style="border:none;border-left:solid blue
                    1.5pt;padding:0in 0in 0in 4.0pt">
                    <div>
                      <div style="border:none;border-top:solid #E1E1E1
                        1.0pt;padding:3.0pt 0in 0in 0in">
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;margin-bottom:12.0pt"><b>From:</b>
                          Martin Duke &lt;<a
                            href="mailto:martin.h.duke@gmail.com"
                            target="_blank" moz-do-not-send="true">martin.h.duke@gmail.com</a>&gt;
                          <br>
                          <b>Sent:</b> Wednesday, June 17, 2020 8:16 PM<o:p></o:p></p>
                      </div>
                    </div>
                    <div>
                      <p class="MsoNormal"
                        style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto">Hi
                        Igor, you might want to check out the "Retry
                        Services" bit of the QUIC-LB draft. This has
                        something to do with the DDoS use case you
                        discuss.<o:p></o:p></p>
                    </div>
                    <p class="MsoNormal"
                      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"> <o:p></o:p></p>
                    <div>
                      <div>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto">On
                          Wed, May 27, 2020 at 9:07 AM Lubashev, Igor
                          &lt;<a href="mailto:ilubashe@akamai.com"
                            target="_blank" moz-do-not-send="true">ilubashe@akamai.com</a>&gt;
                          wrote:<o:p></o:p></p>
                      </div>
                      <blockquote style="border:none;border-left:solid
                        #CCCCCC 1.0pt;padding:0in 0in 0in
6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0in;margin-bottom:5.0pt">
                        <div>
                          <div>
                            <p class="MsoNormal"
                              style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto">I’m
                              working on a manageability draft PR for
                              this (how to rate limit UDP to reduce
                              disruption to QUIC if you have to rate
                              limit UDP).  ETA end of the week (if I do
                              not get pulled into something again).<o:p></o:p></p>
                            <p class="MsoNormal"
                              style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"> <o:p></o:p></p>
                            <p class="MsoNormal"
                              style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto">The
                              relevant observation is that DDoS with UDP
                              that is indistinguishable from QUIC will
                              happen.  UDP is already the most prevalent
                              DDoS vector, since it is easy for a
                              compromised non-admin app to send a flood
                              of huge UDP packets (with TCP you get
                              throttled by the congestion controller). 
                              So there WILL be DDoS protection devices
                              out there to try to mitigate the problem,
                              possibly by observing both directions of
                              the flow and deciding whether a packet
                              belongs to a valid flow or not.<o:p></o:p></p>
                            <p class="MsoNormal"
                              style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"> <o:p></o:p></p>
                            <p class="MsoNormal"
                              style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto">Since
                              such middle boxes will be created, the
                              more explicit and normative Invariants are
                              about what one can expect, the less such
                              middle boxes may decide for themselves. 
                              For example (I did not think long about
                              it), if some elements of path validation
                              could land into Invariants (roughly, “no
                              more than X packets/bytes can be sent on a
                              new path w/o a return packet”), a DDoS
                              middle box may use this fact and active
                              connection migration might still have a
                              chance during an attack (NAT rebinding
                              could be linked by DDoS boxes to an old
                              connection via unchanged CID).<o:p></o:p></p>
                            <p class="MsoNormal"
                              style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"> <o:p></o:p></p>
                            <ul type="disc">
                              <li class="MsoNormal"
                                style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;mso-list:l0
                                level1 lfo1">
                                Igor<o:p></o:p></li>
                            </ul>
                            <p class="MsoNormal"
                              style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"> <o:p></o:p></p>
                            <p class="MsoNormal"
                              style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"> <o:p></o:p></p>
                            <div style="border:none;border-left:solid
                              blue 1.5pt;padding:0in 0in 0in 4.0pt">
                              <p class="MsoNormal"
                                style="mso-margin-top-alt:auto;margin-bottom:12.0pt"><b>From:</b>
                                Christian Huitema &lt;<a
                                  href="mailto:huitema@huitema.net"
                                  target="_blank" moz-do-not-send="true">huitema@huitema.net</a>&gt;
                                <br>
                                <b>Sent:</b> Wednesday, May 27, 2020
                                11:34 AM<o:p></o:p></p>
                              <div>
                                <p class="MsoNormal"
                                  style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto">On
                                  5/27/2020 8:28 AM, Kyle Rose wrote:<o:p></o:p></p>
                              </div>
                              <blockquote
                                style="margin-top:5.0pt;margin-bottom:5.0pt">
                                <p class="MsoNormal"
                                  style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto">On
                                  Wed, May 27, 2020 at 10:34 AM Ian
                                  Swett &lt;ianswett=<a
                                    href="mailto:40google.com@dmarc.ietf.org"
                                    target="_blank"
                                    moz-do-not-send="true">40google.com@dmarc.ietf.org</a>&gt;
                                  wrote:<o:p></o:p></p>
                                <div>
                                  <blockquote
                                    style="border:none;border-left:solid
                                    #CCCCCC 1.0pt;padding:0in 0in 0in
6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0in;margin-bottom:5.0pt">
                                    <div>
                                      <p class="MsoNormal"
                                        style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto">I
                                        was agreeing with MT, but I'm
                                        happy to see some more MUSTs
                                        added if people feel that'd be
                                        helpful.<o:p></o:p></p>
                                    </div>
                                  </blockquote>
                                  <div>
                                    <p class="MsoNormal"
                                      style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
                                        style="font-size:12.0pt"> </span><o:p></o:p></p>
                                    <div>
                                      <div>
                                        <p class="MsoNormal"
                                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
                                            style="font-size:12.0pt">Coincidentally,
                                            we were just talking about
                                            this internally at Akamai
                                            yesterday. IMO, an
                                            invariants document isn't
                                            really helpful if it isn't
                                            normative, and for it to be
                                            normative it (or a related
                                            practices doc for operators)
                                            really needs to spell out
                                            clear boundaries for
                                            operators with MUSTs..</span><o:p></o:p></p>
                                      </div>
                                      <div>
                                        <p class="MsoNormal"
                                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
                                            style="font-size:12.0pt"> </span><o:p></o:p></p>
                                      </div>
                                      <div>
                                        <p class="MsoNormal"
                                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
                                            style="font-size:12.0pt">The
                                            example that came up
                                            yesterday was around
                                            operators filtering QUIC in
                                            the event of a DDoS: one
                                            recommendation based on some
                                            conversations going back at
                                            least to Prague 2019 was to
                                            hash packets on 4-tuple and
                                            filter those below a hash
                                            value chosen for a desired
                                            ingress limit instead of
                                            doing what most operators do
                                            with UDP today, which is to
                                            cap UDP throughput and just
                                            drop packets randomly or
                                            tail drop.</span><o:p></o:p></p>
                                      </div>
                                    </div>
                                  </div>
                                </div>
                              </blockquote>
                              <p>Interesting. Did they consider using
                                the CID, or a fraction of it? This looks
                                entirely like the scenario for which we
                                developed stateless retry.<o:p></o:p></p>
                              <blockquote
                                style="margin-top:5.0pt;margin-bottom:5.0pt">
                                <div>
                                  <div>
                                    <div>
                                      <div>
                                        <p class="MsoNormal"
                                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
                                            style="font-size:12.0pt"> </span><o:p></o:p></p>
                                      </div>
                                      <div>
                                        <p class="MsoNormal"
                                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
                                            style="font-size:12.0pt">This
                                            recommendation certainly
                                            imposes some constraints on
                                            future protocol development
                                            that motivate new
                                            invariants: for instance, it
                                            would preclude sharding a
                                            connection across multiple
                                            source ports (not that there
                                            is necessarily a reason to
                                            do this; it's just an
                                            example). But more
                                            importantly, it goes beyond
                                            invariants: it's one among
                                            many practices compatible
                                            with the current set of
                                            invariants, some reasonable
                                            and some terrible.</span><o:p></o:p></p>
                                      </div>
                                    </div>
                                  </div>
                                </div>
                              </blockquote>
                              <p>This would break the "preferred
                                address" redirection. Preferred address
                                migration may or may not be spelled out
                                in the invariants.<o:p></o:p></p>
                              <blockquote
                                style="margin-top:5.0pt;margin-bottom:5.0pt">
                                <div>
                                  <div>
                                    <div>
                                      <div>
                                        <p class="MsoNormal"
                                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
                                            style="font-size:12.0pt"> </span><o:p></o:p></p>
                                      </div>
                                      <div>
                                        <p class="MsoNormal"
                                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
                                            style="font-size:12.0pt">Operators
                                            are going to do things to
                                            QUIC traffic, so it would be
                                            good to offer them
                                            recommendations that are
                                            compatible with broad
                                            deployability.</span><o:p></o:p></p>
                                      </div>
                                    </div>
                                  </div>
                                </div>
                              </blockquote>
                              <p> <o:p></o:p></p>
                              <p>Yes, we do need the invariants for
                                that.<o:p></o:p></p>
                              <p>-- Christian Huitema<o:p></o:p></p>
                            </div>
                          </div>
                        </div>
                      </blockquote>
                    </div>
                  </div>
                </div>
              </div>
            </blockquote>
          </div>
        </div>
      </div>
    </blockquote>
  </body>
</html>

--------------78BEF07D38E0F11540D1141A--


From nobody Fri Jun 19 14:53:40 2020
Return-Path: <bkaduk@akamai.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 31CBC3A0EB5 for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 14:53:39 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.1
X-Spam-Level: 
X-Spam-Status: No, score=-2.1 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=unavailable autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=akamai.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id R4PziQzrZj0g for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 14:53:38 -0700 (PDT)
Received: from mx0a-00190b01.pphosted.com (mx0a-00190b01.pphosted.com [IPv6:2620:100:9001:583::1]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 053303A0EB3 for <quic@ietf.org>; Fri, 19 Jun 2020 14:53:37 -0700 (PDT)
Received: from pps.filterd (m0122332.ppops.net [127.0.0.1]) by mx0a-00190b01.pphosted.com (8.16.0.42/8.16.0.42) with SMTP id 05JLmkQJ004453; Fri, 19 Jun 2020 22:53:29 +0100
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=akamai.com; h=date : from : to : cc : subject : message-id : references : mime-version : content-type : in-reply-to; s=jan2016.eng; bh=anecQlIMAEQB3rlazaaizObML9FpjzsLrjdGoFMfo1s=; b=lpvSJep+K7610x4FTNIJnZiKe4Nv2V0ZhTnuWlc/f3qoCkU8DcfGhVxBMe0TYxItmkUa ar31//nmAHLUpQdSzS4BR/dHRkvph+7lponn5Il2/BfXnJjAaCVRzeC9Kr+zphFu6Cv8 dN/x7S1uaCauLSZSN/09OYfl4h+Xk9i7M0tJ/b74GcfeYwUGWQZ/b3K7eV+pHo55+3mg aufCsIO/Xk29TwE9tnHYg/ZsuJ4EqxN6MFGgwXfNXd6KLhjDLXuzXnAYn7SKMfKXxDmm RXlo2HzZBEWyWOmkchwKtCuGpPU20YQ6CH1WTaJ6+c8NB1PNdjikur8tDEiFQePBSDA6 Uw== 
Received: from prod-mail-ppoint2 (prod-mail-ppoint2.akamai.com [184.51.33.19] (may be forged)) by mx0a-00190b01.pphosted.com with ESMTP id 31qqyc961s-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Fri, 19 Jun 2020 22:53:29 +0100
Received: from pps.filterd (prod-mail-ppoint2.akamai.com [127.0.0.1]) by prod-mail-ppoint2.akamai.com (8.16.0.42/8.16.0.42) with SMTP id 05JLkrjp003631; Fri, 19 Jun 2020 17:53:28 -0400
Received: from prod-mail-relay18.dfw02.corp.akamai.com ([172.27.165.172]) by prod-mail-ppoint2.akamai.com with ESMTP id 31qjm1eac1-1; Fri, 19 Jun 2020 17:53:28 -0400
Received: from akamai.com (unknown [172.19.16.134]) by prod-mail-relay18.dfw02.corp.akamai.com (Postfix) with ESMTP id 46D86885; Fri, 19 Jun 2020 21:53:26 +0000 (GMT)
Date: Fri, 19 Jun 2020 14:53:25 -0700
From: Benjamin Kaduk <bkaduk@akamai.com>
To: Christian Huitema <huitema@huitema.net>
Cc: "Lubashev, Igor" <ilubashe=40akamai.com@dmarc.ietf.org>, Martin Duke <martin.h.duke@gmail.com>, IETF QUIC WG <quic@ietf.org>, Lars Eggert <lars@eggert.org>, Kyle Rose <krose@krose.org>, Ian Swett <ianswett=40google.com@dmarc.ietf.org>, Jared Mauch <jared@puck.nether.net>
Subject: Re: Is the invariants draft really standards track?
Message-ID: <20200619215324.GF20623@akamai.com>
References: <833A693C-62E6-4889-9954-FCE65A839A7C@eggert.org> <CAKcm_gPMO2DtqvKucqVw0zDjSniSOmFD4p1Tp4YLjr9WSWdEUw@mail.gmail.com> <CAJU8_nUN42gGmQof24XD9-EjXedyzcarDyRP8MGe1qW-BZ=+Aw@mail.gmail.com> <9cd91c24-c730-22a4-7aa0-baf61613b3ce@huitema.net> <f4922cdb59014202900de44cc5fea0ff@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQvwkTvpUcu6-+W5zWo22m-R1jvN7DcCpXfuw8Hb55qsw@mail.gmail.com> <95dd02c92b32472d9cab0dd47b98c637@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQxxXn27rZEY75-AsHD5VF0fqiV1VDyeSrzQ=-sM7JNCg@mail.gmail.com> <9c2e300c30f74d1794d11cf4334ea07b@usma1ex-dag1mb5.msg.corp.akamai.com> <2c40f3d9-fa40-9834-ac30-36bc9a1a6303@huitema.net>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <2c40f3d9-fa40-9834-ac30-36bc9a1a6303@huitema.net>
User-Agent: Mutt/1.9.4 (2018-02-28)
X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.216, 18.0.687 definitions=2020-06-19_22:2020-06-19, 2020-06-19 signatures=0
X-Proofpoint-Spam-Details: rule=notspam policy=default score=0 spamscore=0 bulkscore=0 suspectscore=0 phishscore=0 mlxlogscore=394 adultscore=0 mlxscore=0 malwarescore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2004280000 definitions=main-2006190151
X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.216, 18.0.687 definitions=2020-06-19_22:2020-06-19, 2020-06-19 signatures=0
X-Proofpoint-Spam-Details: rule=notspam policy=default score=0 spamscore=0 malwarescore=0 bulkscore=0 lowpriorityscore=0 impostorscore=0 priorityscore=1501 phishscore=0 suspectscore=0 mlxlogscore=327 mlxscore=0 clxscore=1011 adultscore=0 cotscore=-2147483648 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2004280000 definitions=main-2006190152
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/QhWjIMMoqzTWiOJJ_-zACtPj-JE>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 19 Jun 2020 21:53:39 -0000

On Fri, Jun 19, 2020 at 02:48:48PM -0700, Christian Huitema wrote:
> When under DOS attack, you want to "minimize blowback", i.e., as much as
> possible avoid generating packets in response to attack traffic. So,
> yes, a server may choose to not send stateless resets to anyone when
> under attack; in fact, my recommendation would be that a server SHOULD
> NOT send stateless resets to anyone when under attack.
> 
> That said, Igor raised an interesting point about return traffic. It
> would be very nice if DOS protection boxes could distinguish between
> "validated traffic" that the server presumably intends to process, and
> "unsolicited traffic" that will just consume resource. The box can then
> reserve some share of the resource for validated traffic, and place the
> rest of the traffic in a lower priority queue. Fine, but there needs to
> be a test. The classic test is that incoming traffic is "validated" if
> the protection box can match it with return traffic coming from the
> server -- for some definition of matching.
> 
> From that point of view, stateless reset is definitely not helpful. But
> problematic traffic goes beyond that. The server will reply to a
> client's initial with a server's initial packet. Does that validate the
> response traffic? OK, maybe the protection box can programmed to only
> validate traffic if it sees 1RTT packets. But many servers will send
> 1-RTT packets as 0.5 RTT. Does that validate the response traffic?
> 
> We might say that traffic is validated when the handshake is confirmed,
> but the protection box does not understand the TLS handshake, it just
> sees packet types and packet sizes. It cannot distinguish between 0.5RTT
> data and 1RTT data, and thus the closest approximation of "validation"
> would be seeing more than an initial window worth of traffic coming back
> from the server. That does not sound great.
> 
> On the other hand, things get much better if the server under attack can
> adopt a defensive posture and help the DOS protection box do its job.
> Suppose that a server can detect that it is under attack -- or be
> explicitly configured so. The simplest defensive posture would be to (1)
> disable stateless reset and (2) not send any 0.5RTT packet, including in
> response to 0-RTT. The protection boxes can at that point take the 1-RTT
> packets from the server as indicating validation.
> 
> Maybe we should specify that.

It might also be an interesting thought experiment to see what
kind of things the QUIC server could send to the DoS protection box if
the QUIC server acts as a DOTS (RFCs 8782 and 8783) client to a DOTS
server controlling the DoS protection box.

-Ben


From nobody Fri Jun 19 15:05:51 2020
Return-Path: <ted.ietf@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id E6C043A0EE3 for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 15:05:44 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.997
X-Spam-Level: 
X-Spam-Status: No, score=-1.997 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, HTTPS_HTTP_MISMATCH=0.1, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id NzZ_Q2RTkHjr for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 15:05:42 -0700 (PDT)
Received: from mail-ot1-x329.google.com (mail-ot1-x329.google.com [IPv6:2607:f8b0:4864:20::329]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 5B63A3A0F1E for <quic@ietf.org>; Fri, 19 Jun 2020 15:05:40 -0700 (PDT)
Received: by mail-ot1-x329.google.com with SMTP id n5so8484405otj.1 for <quic@ietf.org>; Fri, 19 Jun 2020 15:05:40 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=7zAdIQYgmCoQDPGwx921lv5h/Yn261qQ8dODQnrHg4E=; b=dBrBhOJCihOnuOtqyh0y2ak1Yvq/5zrX4BJ3RDCUJgCvC8m52MJpHlOYMbPEdWisF3 0rDfzXhSh05eIiw7fegunIBoblFUIeeAk+WM+H20m8x42y5gvYHcwvczAil4/HTSWU9M AVSAGXUUztT7bXlrYW9+9JUTguNkHWL/6aINkz5MMNdo0F1Qvl88huLrnlB0ZsS0zF4/ 89AdlCk78UCg8jWjDfmXQayb1htN3Mw0s4+5/By5b4aMQw9CZ6wbtiFIhNNeTP55ShYT TWrJoUMjjVFZEDd6E3qcxk2/FTXFTaWhs+cFScfDvyU/2KPPyaob8QWVqqQy0OFN8M2e JOKg==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=7zAdIQYgmCoQDPGwx921lv5h/Yn261qQ8dODQnrHg4E=; b=Vyt73q6VuBSuwJJ5GvXmTdGwOZ9TgS/WG0ypAQb+W0JEhwGS2E376W5miEY2W+Y7j1 rS8b8jwE8KR8FZqPolhsTPXxEfQQ6GHHYV3uKaAKIoZwAZe64mPVmh2d+R9vJp8MiAdD B70tGTZa/DMxlFuhz3YMEqLJSjB9UJQ3V6eww+jdumR6AZRf9UsSdRkubhWT7A20Hvyl qENvzg4hgRa2chLdS5ENHbCd1MnE5T3FP2mBP1169LUgSYuTWVfqYPRfwD2CJM+m7/d4 gjgliRRTLBfG5kcuPTKeqFjP6zMH/leP0uJd/DFOCGslQRlcCuoaz/YBg973Nhbu4P8C vHxw==
X-Gm-Message-State: AOAM532vJM/P6TCsLIU7V9OtsyGbL3PV3JUk5tPJTKrmHA2MwuElt0Gn i52xXKoa+axqhTUvKH9H6TlrCBT3V8DYB+m96Ss=
X-Google-Smtp-Source: ABdhPJxB6Zk9Lt4xn4imCkRITXCnZq9PCpffoaWBMF7JiS8zUg/Ic3wAPbtfXoPG9s3BcR5WefxwrzCqDKdEPhMtM0o=
X-Received: by 2002:a9d:aaa:: with SMTP id 39mr4555472otq.269.1592604339577; Fri, 19 Jun 2020 15:05:39 -0700 (PDT)
MIME-Version: 1.0
References: <CAM4esxQBqfrz24riPQA_VGKcGp_TzW0pqb97KfFMtNdW9pUfDg@mail.gmail.com> <833A693C-62E6-4889-9954-FCE65A839A7C@eggert.org> <CAKcm_gPMO2DtqvKucqVw0zDjSniSOmFD4p1Tp4YLjr9WSWdEUw@mail.gmail.com> <CAJU8_nUN42gGmQof24XD9-EjXedyzcarDyRP8MGe1qW-BZ=+Aw@mail.gmail.com> <9cd91c24-c730-22a4-7aa0-baf61613b3ce@huitema.net> <f4922cdb59014202900de44cc5fea0ff@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQvwkTvpUcu6-+W5zWo22m-R1jvN7DcCpXfuw8Hb55qsw@mail.gmail.com> <95dd02c92b32472d9cab0dd47b98c637@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQxxXn27rZEY75-AsHD5VF0fqiV1VDyeSrzQ=-sM7JNCg@mail.gmail.com> <9c2e300c30f74d1794d11cf4334ea07b@usma1ex-dag1mb5.msg.corp.akamai.com> <2c40f3d9-fa40-9834-ac30-36bc9a1a6303@huitema.net>
In-Reply-To: <2c40f3d9-fa40-9834-ac30-36bc9a1a6303@huitema.net>
From: Ted Hardie <ted.ietf@gmail.com>
Date: Fri, 19 Jun 2020 15:05:13 -0700
Message-ID: <CA+9kkMBQt001xOVgT=9G8YOOTOJ+9S=OWDwGEeYuKVm46Cq3iQ@mail.gmail.com>
Subject: Re: Is the invariants draft really standards track?
To: Christian Huitema <huitema@huitema.net>
Cc: "Lubashev, Igor" <ilubashe=40akamai.com@dmarc.ietf.org>,  Martin Duke <martin.h.duke@gmail.com>, IETF QUIC WG <quic@ietf.org>, Lars Eggert <lars@eggert.org>,  Kyle Rose <krose@krose.org>, Ian Swett <ianswett=40google.com@dmarc.ietf.org>,  Jared Mauch <jared@puck.nether.net>
Content-Type: multipart/alternative; boundary="000000000000edac6605a8771764"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/Vpmve31uAjadB3AbCN4WHfJ1dzE>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 19 Jun 2020 22:05:49 -0000

--000000000000edac6605a8771764
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Hi Christian,

A question in-line,

On Fri, Jun 19, 2020 at 2:49 PM Christian Huitema <huitema@huitema.net>
wrote:

> When under DOS attack, you want to "minimize blowback", i.e., as much as
> possible avoid generating packets in response to attack traffic. So, yes,=
 a
> server may choose to not send stateless resets to anyone when under attac=
k;
> in fact, my recommendation would be that a server SHOULD NOT send statele=
ss
> resets to anyone when under attack.
>
> That said, Igor raised an interesting point about return traffic. It woul=
d
> be very nice if DOS protection boxes could distinguish between "validated
> traffic" that the server presumably intends to process, and "unsolicited
> traffic" that will just consume resource. The box can then reserve some
> share of the resource for validated traffic, and place the rest of the
> traffic in a lower priority queue. Fine, but there needs to be a test. Th=
e
> classic test is that incoming traffic is "validated" if the protection bo=
x
> can match it with return traffic coming from the server -- for some
> definition of matching.
>
> From that point of view, stateless reset is definitely not helpful. But
> problematic traffic goes beyond that. The server will reply to a client's
> initial with a server's initial packet. Does that validate the response
> traffic? OK, maybe the protection box can programmed to only validate
> traffic if it sees 1RTT packets. But many servers will send 1-RTT packets
> as 0.5 RTT. Does that validate the response traffic?
>
> We might say that traffic is validated when the handshake is confirmed,
> but the protection box does not understand the TLS handshake, it just see=
s
> packet types and packet sizes. It cannot distinguish between 0.5RTT data
> and 1RTT data, and thus the closest approximation of "validation" would b=
e
> seeing more than an initial window worth of traffic coming back from the
> server. That does not sound great.
>
> On the other hand, things get much better if the server under attack can
> adopt a defensive posture and help the DOS protection box do its job.
> Suppose that a server can detect that it is under attack -- or be
> explicitly configured so. The simplest defensive posture would be to (1)
> disable stateless reset and (2) not send any 0.5RTT packet, including in
> response to 0-RTT. The protection boxes can at that point take the 1-RTT
> packets from the server as indicating validation.
>
Perhaps I'm misreading you here, but this sounds like you expect the
protection boxes to be able to distinguish when servers see themselves as
under DOS from when they don't, so that they can tell that the lack of
0.5RTT is an indication of an attack response.  Given distribution patterns
of DOS attacks, I'm struggling to see whether that will commonly be the
case.

You could, of course, always put handshake traffic into low-priority queues
until you see the 1-RTT packets that validate the server's interest.  That
would make 1-RTT traffic effectively a path signal of validation.  My
concern with that is that using those low priority queues during the
handshake phase seems likely to result in worse latency and increased risk
of packet loss (which can be tricky during that phase).  That seems a heavy
price to pay during non-attack times for better protection when under
attack.

Am I misunderstanding how this is distinguished?

Clue appreciated,

Ted



> Maybe we should specify that.
>
> -- Christian Huitema
> On 6/19/2020 1:42 PM, Lubashev, Igor wrote:
>
> > There is no need for servers to decrypt CIDs in QUIC-LB. Presumably the
> server has a lookup table for its CIDs.
>
>
>
> Sending a stateless reset in response to a junk packet would cost more CP=
U
> than verifying CID integrity.  But, yes, a server may choose to not send
> stateless resets to anyone when under attack.
>
>
>
>
>
> *From:* Martin Duke <martin.h.duke@gmail.com> <martin.h.duke@gmail.com>
> *Sent:* Friday, June 19, 2020 2:44 PM
>
> >Unfortunately, Retry system protects only server's memory state and some
> CPU cycles spent on crypto.  (Servers still need to decrypt CID to decide
> it is invalid, and if the attacker is clever enough to establish one vali=
d
> connection and use that CID in a flood, the server will also be decryptin=
g
> packets.)
>
>
>
> There is no need for servers to decrypt CIDs in QUIC-LB. Presumably the
> server has a lookup table for its CIDs.
>
>
>
> It is true that Retry Services (and indeed, the Retry concept as a whole)
> does nothing to protect network capacity.
>
>
>
> On Fri, Jun 19, 2020 at 8:08 AM Lubashev, Igor <ilubashe@akamai.com>
> wrote:
>
> It looks like
> https://tools.ietf.org/html/draft-ietf-quic-load-balancers-02#section-5
> <https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__tools.ietf.org_ht=
ml_draft-2Dietf-2Dquic-2Dload-2Dbalancers-2D02-23section-2D5&d=3DDwMFaQ&c=
=3D96ZbZZcaMF4w0F4jpN6LZg&r=3DDjn3bQ5uNJDPM_2skfL3rW1tzcIxyjUZdn_m55KPmlo&m=
=3DeTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTKljvj-ys&s=3D1x7AN2a51X_zV5xSyK0uCL8v=
Z5cugD3n4lbFWZDqVQg&e=3D>
> is an excellent discussion of Retry mechanics.  It definitely deserves a
> reference from this manageability draft.
>
>
>
> The Retry mechanisms described in LB draft are all cooperating boxes, and
> servers must be aware of them.  Unfortunately, Retry system protects only
> server's memory state and some CPU cycles spent on crypto.  (Servers stil=
l
> need to decrypt CID to decide it is invalid, and if the attacker is cleve=
r
> enough to establish one valid connection and use that CID in a flood, the
> server will also be decrypting packets.)  Retry does nothing to protect
> network resources.
>
>
>
> The PR I opened (https://github.com/quicwg/ops-drafts/pull/94
> <https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__github.com_quicwg=
_ops-2Ddrafts_pull_94&d=3DDwMFaQ&c=3D96ZbZZcaMF4w0F4jpN6LZg&r=3DDjn3bQ5uNJD=
PM_2skfL3rW1tzcIxyjUZdn_m55KPmlo&m=3DeTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTKlj=
vj-ys&s=3DVpvhQ9n79rANIk3O5Sm7PZyToFQEennoPt9iqFpWbq8&e=3D>)
> is about uncooperating devices that try to mitigate volumetric network
> attacks.
>
>
>
>
>
> *From:* Martin Duke <martin.h.duke@gmail.com>
> *Sent:* Wednesday, June 17, 2020 8:16 PM
>
> Hi Igor, you might want to check out the "Retry Services" bit of the
> QUIC-LB draft. This has something to do with the DDoS use case you discus=
s.
>
>
>
> On Wed, May 27, 2020 at 9:07 AM Lubashev, Igor <ilubashe@akamai.com>
> wrote:
>
> I=E2=80=99m working on a manageability draft PR for this (how to rate lim=
it UDP to
> reduce disruption to QUIC if you have to rate limit UDP).  ETA end of the
> week (if I do not get pulled into something again).
>
>
>
> The relevant observation is that DDoS with UDP that is indistinguishable
> from QUIC will happen.  UDP is already the most prevalent DDoS vector,
> since it is easy for a compromised non-admin app to send a flood of huge
> UDP packets (with TCP you get throttled by the congestion controller).  S=
o
> there WILL be DDoS protection devices out there to try to mitigate the
> problem, possibly by observing both directions of the flow and deciding
> whether a packet belongs to a valid flow or not.
>
>
>
> Since such middle boxes will be created, the more explicit and normative
> Invariants are about what one can expect, the less such middle boxes may
> decide for themselves.  For example (I did not think long about it), if
> some elements of path validation could land into Invariants (roughly, =E2=
=80=9Cno
> more than X packets/bytes can be sent on a new path w/o a return packet=
=E2=80=9D),
> a DDoS middle box may use this fact and active connection migration might
> still have a chance during an attack (NAT rebinding could be linked by DD=
oS
> boxes to an old connection via unchanged CID).
>
>
>
>    - Igor
>
>
>
>
>
> *From:* Christian Huitema <huitema@huitema.net>
> *Sent:* Wednesday, May 27, 2020 11:34 AM
>
> On 5/27/2020 8:28 AM, Kyle Rose wrote:
>
> On Wed, May 27, 2020 at 10:34 AM Ian Swett <ianswett=3D
> 40google.com@dmarc.ietf.org> wrote:
>
> I was agreeing with MT, but I'm happy to see some more MUSTs added if
> people feel that'd be helpful.
>
>
>
> Coincidentally, we were just talking about this internally at Akamai
> yesterday. IMO, an invariants document isn't really helpful if it isn't
> normative, and for it to be normative it (or a related practices doc for
> operators) really needs to spell out clear boundaries for operators with
> MUSTs..
>
>
>
> The example that came up yesterday was around operators filtering QUIC in
> the event of a DDoS: one recommendation based on some conversations going
> back at least to Prague 2019 was to hash packets on 4-tuple and filter
> those below a hash value chosen for a desired ingress limit instead of
> doing what most operators do with UDP today, which is to cap UDP throughp=
ut
> and just drop packets randomly or tail drop.
>
> Interesting. Did they consider using the CID, or a fraction of it? This
> looks entirely like the scenario for which we developed stateless retry.
>
>
>
> This recommendation certainly imposes some constraints on future protocol
> development that motivate new invariants: for instance, it would preclude
> sharding a connection across multiple source ports (not that there is
> necessarily a reason to do this; it's just an example). But more
> importantly, it goes beyond invariants: it's one among many practices
> compatible with the current set of invariants, some reasonable and some
> terrible.
>
> This would break the "preferred address" redirection. Preferred address
> migration may or may not be spelled out in the invariants.
>
>
>
> Operators are going to do things to QUIC traffic, so it would be good to
> offer them recommendations that are compatible with broad deployability.
>
>
>
> Yes, we do need the invariants for that.
>
> -- Christian Huitema
>
>

--000000000000edac6605a8771764
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div dir=3D"ltr"><div>Hi Christian,</div><div><br></div><d=
iv>A question in-line,<br></div></div><br><div class=3D"gmail_quote"><div d=
ir=3D"ltr" class=3D"gmail_attr">On Fri, Jun 19, 2020 at 2:49 PM Christian H=
uitema &lt;<a href=3D"mailto:huitema@huitema.net">huitema@huitema.net</a>&g=
t; wrote:<br></div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0p=
x 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
 =20
   =20
 =20
  <div>
    <p>When under DOS attack, you want to &quot;minimize blowback&quot;, i.=
e., as
      much as possible avoid generating packets in response to attack
      traffic. So, yes, a server may choose to not send stateless resets
      to anyone when under attack; in fact, my recommendation would be
      that a server SHOULD NOT send stateless resets to anyone when
      under attack.</p>
    <p>That said, Igor raised an interesting point about return traffic.
      It would be very nice if DOS protection boxes could distinguish
      between &quot;validated traffic&quot; that the server presumably inte=
nds to
      process, and &quot;unsolicited traffic&quot; that will just consume
      resource. The box can then reserve some share of the resource for
      validated traffic, and place the rest of the traffic in a lower
      priority queue. Fine, but there needs to be a test. The classic
      test is that incoming traffic is &quot;validated&quot; if the protect=
ion box
      can match it with return traffic coming from the server -- for
      some definition of matching.</p>
    <p>From that point of view, stateless reset is definitely not
      helpful. But problematic traffic goes beyond that. The server will
      reply to a client&#39;s initial with a server&#39;s initial packet. D=
oes
      that validate the response traffic? OK, maybe the protection box
      can programmed to only validate traffic if it sees 1RTT packets.
      But many servers will send 1-RTT packets as 0.5 RTT. Does that
      validate the response traffic?</p>
    <p>We might say that traffic is validated when the handshake is
      confirmed, but the protection box does not understand the TLS
      handshake, it just sees packet types and packet sizes. It cannot
      distinguish between 0.5RTT data and 1RTT data, and thus the
      closest approximation of &quot;validation&quot; would be seeing more =
than an
      initial window worth of traffic coming back from the server. That
      does not sound great.</p>
    <p>On the other hand, things get much better if the server under
      attack can adopt a defensive posture and help the DOS protection
      box do its job. Suppose that a server can detect that it is under
      attack -- or be explicitly configured so. The simplest defensive
      posture would be to (1) disable stateless reset and (2) not send
      any 0.5RTT packet, including in response to 0-RTT. The protection
      boxes can at that point take the 1-RTT packets from the server as
      indicating validation.</p></div></blockquote><div>Perhaps I&#39;m mis=
reading you here, but this sounds like you expect the protection boxes to b=
e able to distinguish when servers see themselves as under DOS from when th=
ey don&#39;t, so that they can tell that the lack of 0.5RTT is an indicatio=
n of an attack response.=C2=A0 Given distribution patterns of DOS attacks, =
I&#39;m struggling to see whether that will commonly be the case.</div><div=
><br></div><div>You could, of course, always put handshake traffic into low=
-priority queues until you see the 1-RTT packets that validate the server&#=
39;s interest.=C2=A0 That would make 1-RTT traffic effectively a path signa=
l of validation.=C2=A0 My concern with that is that using those low priorit=
y queues during the handshake phase seems likely to result in worse latency=
 and increased risk of packet loss (which can be tricky during that phase).=
=C2=A0 That seems a heavy price to pay during non-attack times for better p=
rotection when under attack.</div><div><br></div><div>Am I misunderstanding=
 how this is distinguished?</div><div><br></div><div>Clue appreciated,</div=
><div><br></div><div>Ted<br></div><div><br></div><div>=C2=A0</div><blockquo=
te class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px =
solid rgb(204,204,204);padding-left:1ex"><div>
    <p>Maybe we should specify that.</p>
    <p>-- Christian Huitema<br>
    </p>
    <div>On 6/19/2020 1:42 PM, Lubashev, Igor
      wrote:<br>
    </div>
    <blockquote type=3D"cite">
     =20
     =20
     =20
      <div>
        <p class=3D"MsoNormal">&gt; There is no need for servers to
          decrypt CIDs in QUIC-LB. Presumably the server has a lookup
          table for its CIDs.<u></u><u></u></p>
        <p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
        <p class=3D"MsoNormal">Sending a stateless reset in response to a
          junk packet would cost more CPU than verifying CID integrity.=C2=
=A0
          But, yes, a server may choose to not send stateless resets to
          anyone when under attack.<u></u><u></u></p>
        <p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
        <p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
        <div style=3D"border-color:currentcolor currentcolor currentcolor b=
lue;border-style:none none none solid;border-width:medium medium medium 1.5=
pt;padding:0in 0in 0in 4pt">
          <div>
            <div style=3D"border-color:rgb(225,225,225) currentcolor curren=
tcolor;border-style:solid none none;border-width:1pt medium medium;padding:=
3pt 0in 0in">
              <p class=3D"MsoNormal"><b>From:</b> Martin Duke
                <a href=3D"mailto:martin.h.duke@gmail.com" target=3D"_blank=
">&lt;martin.h.duke@gmail.com&gt;</a> <br>
                <b>Sent:</b> Friday, June 19, 2020 2:44 PM<br>
                <br>
                <u></u><u></u></p>
            </div>
          </div>
          <div>
            <p class=3D"MsoNormal">&gt;Unfortunately, Retry system
              protects only server&#39;s memory state and some CPU cycles
              spent on crypto.=C2=A0 (Servers still need to decrypt CID to
              decide it is invalid, and if the attacker is clever enough
              to establish one valid connection and use that CID in a
              flood, the server will also be decrypting packets.)=C2=A0=C2=
=A0<u></u><u></u></p>
            <div>
              <p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
            </div>
            <div>
              <p class=3D"MsoNormal">There is no need for servers to
                decrypt CIDs in QUIC-LB. Presumably the server has a
                lookup table for its CIDs.<u></u><u></u></p>
            </div>
            <div>
              <p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
            </div>
            <div>
              <p class=3D"MsoNormal">It is true that Retry Services (and
                indeed, the Retry concept as a whole) does nothing to
                protect network capacity.<u></u><u></u></p>
            </div>
          </div>
          <p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
          <div>
            <div>
              <p class=3D"MsoNormal">On Fri, Jun 19, 2020 at 8:08 AM
                Lubashev, Igor &lt;<a href=3D"mailto:ilubashe@akamai.com" t=
arget=3D"_blank">ilubashe@akamai.com</a>&gt;
                wrote:<u></u><u></u></p>
            </div>
            <blockquote style=3D"border-color:currentcolor currentcolor cur=
rentcolor rgb(204,204,204);border-style:none none none solid;border-width:m=
edium medium medium 1pt;padding:0in 0in 0in 6pt;margin-left:4.8pt;margin-ri=
ght:0in">
              <div>
                <div>
                  <p class=3D"MsoNormal">It
                    looks like
                    <a href=3D"https://urldefense.proofpoint.com/v2/url?u=
=3Dhttps-3A__tools.ietf.org_html_draft-2Dietf-2Dquic-2Dload-2Dbalancers-2D0=
2-23section-2D5&amp;d=3DDwMFaQ&amp;c=3D96ZbZZcaMF4w0F4jpN6LZg&amp;r=3DDjn3b=
Q5uNJDPM_2skfL3rW1tzcIxyjUZdn_m55KPmlo&amp;m=3DeTT-BoZ1fMitywKRVSxpU3js0lhO=
0qkspYTKljvj-ys&amp;s=3D1x7AN2a51X_zV5xSyK0uCL8vZ5cugD3n4lbFWZDqVQg&amp;e=
=3D" target=3D"_blank">
https://tools.ietf.org/html/draft-ietf-quic-load-balancers-02#section-5</a>
                    is an excellent discussion of Retry mechanics.=C2=A0 It
                    definitely deserves a reference from this
                    manageability draft.<u></u><u></u></p>
                  <p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
                  <p class=3D"MsoNormal">The
                    Retry mechanisms described in LB draft are all
                    cooperating boxes, and servers must be aware of
                    them.=C2=A0 Unfortunately, Retry system protects only
                    server&#39;s memory state and some CPU cycles spent on
                    crypto.=C2=A0 (Servers still need to decrypt CID to
                    decide it is invalid, and if the attacker is clever
                    enough to establish one valid connection and use
                    that CID in a flood, the server will also be
                    decrypting packets.)=C2=A0 Retry does nothing to protec=
t
                    network resources.<u></u><u></u></p>
                  <p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
                  <p class=3D"MsoNormal">The
                    PR I opened (<a href=3D"https://urldefense.proofpoint.c=
om/v2/url?u=3Dhttps-3A__github.com_quicwg_ops-2Ddrafts_pull_94&amp;d=3DDwMF=
aQ&amp;c=3D96ZbZZcaMF4w0F4jpN6LZg&amp;r=3DDjn3bQ5uNJDPM_2skfL3rW1tzcIxyjUZd=
n_m55KPmlo&amp;m=3DeTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTKljvj-ys&amp;s=3DVpvh=
Q9n79rANIk3O5Sm7PZyToFQEennoPt9iqFpWbq8&amp;e=3D" target=3D"_blank">https:/=
/github.com/quicwg/ops-drafts/pull/94</a>)
                    is about uncooperating devices that try to mitigate
                    volumetric network attacks.<u></u><u></u></p>
                  <p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
                  <p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
                  <div style=3D"border-color:currentcolor currentcolor curr=
entcolor blue;border-style:none none none solid;border-width:medium medium =
medium 1.5pt;padding:0in 0in 0in 4pt">
                    <div>
                      <div style=3D"border-color:rgb(225,225,225) currentco=
lor currentcolor;border-style:solid none none;border-width:1pt medium mediu=
m;padding:3pt 0in 0in">
                        <p class=3D"MsoNormal" style=3D"margin-bottom:12pt"=
><b>From:</b>
                          Martin Duke &lt;<a href=3D"mailto:martin.h.duke@g=
mail.com" target=3D"_blank">martin.h.duke@gmail.com</a>&gt;
                          <br>
                          <b>Sent:</b> Wednesday, June 17, 2020 8:16 PM<u><=
/u><u></u></p>
                      </div>
                    </div>
                    <div>
                      <p class=3D"MsoNormal">Hi
                        Igor, you might want to check out the &quot;Retry
                        Services&quot; bit of the QUIC-LB draft. This has
                        something to do with the DDoS use case you
                        discuss.<u></u><u></u></p>
                    </div>
                    <p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
                    <div>
                      <div>
                        <p class=3D"MsoNormal">On
                          Wed, May 27, 2020 at 9:07 AM Lubashev, Igor
                          &lt;<a href=3D"mailto:ilubashe@akamai.com" target=
=3D"_blank">ilubashe@akamai.com</a>&gt;
                          wrote:<u></u><u></u></p>
                      </div>
                      <blockquote style=3D"border-color:currentcolor curren=
tcolor currentcolor rgb(204,204,204);border-style:none none none solid;bord=
er-width:medium medium medium 1pt;padding:0in 0in 0in 6pt;margin:5pt 0in 5p=
t 4.8pt">
                        <div>
                          <div>
                            <p class=3D"MsoNormal">I=E2=80=99m
                              working on a manageability draft PR for
                              this (how to rate limit UDP to reduce
                              disruption to QUIC if you have to rate
                              limit UDP).=C2=A0 ETA end of the week (if I d=
o
                              not get pulled into something again).<u></u><=
u></u></p>
                            <p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
                            <p class=3D"MsoNormal">The
                              relevant observation is that DDoS with UDP
                              that is indistinguishable from QUIC will
                              happen.=C2=A0 UDP is already the most prevale=
nt
                              DDoS vector, since it is easy for a
                              compromised non-admin app to send a flood
                              of huge UDP packets (with TCP you get
                              throttled by the congestion controller).=C2=
=A0
                              So there WILL be DDoS protection devices
                              out there to try to mitigate the problem,
                              possibly by observing both directions of
                              the flow and deciding whether a packet
                              belongs to a valid flow or not.<u></u><u></u>=
</p>
                            <p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
                            <p class=3D"MsoNormal">Since
                              such middle boxes will be created, the
                              more explicit and normative Invariants are
                              about what one can expect, the less such
                              middle boxes may decide for themselves.=C2=A0
                              For example (I did not think long about
                              it), if some elements of path validation
                              could land into Invariants (roughly, =E2=80=
=9Cno
                              more than X packets/bytes can be sent on a
                              new path w/o a return packet=E2=80=9D), a DDo=
S
                              middle box may use this fact and active
                              connection migration might still have a
                              chance during an attack (NAT rebinding
                              could be linked by DDoS boxes to an old
                              connection via unchanged CID).<u></u><u></u><=
/p>
                            <p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
                            <ul type=3D"disc">
                              <li class=3D"MsoNormal">
                                Igor<u></u><u></u></li>
                            </ul>
                            <p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
                            <p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
                            <div style=3D"border-color:currentcolor current=
color currentcolor blue;border-style:none none none solid;border-width:medi=
um medium medium 1.5pt;padding:0in 0in 0in 4pt">
                              <p class=3D"MsoNormal" style=3D"margin-bottom=
:12pt"><b>From:</b>
                                Christian Huitema &lt;<a href=3D"mailto:hui=
tema@huitema.net" target=3D"_blank">huitema@huitema.net</a>&gt;
                                <br>
                                <b>Sent:</b> Wednesday, May 27, 2020
                                11:34 AM<u></u><u></u></p>
                              <div>
                                <p class=3D"MsoNormal">On
                                  5/27/2020 8:28 AM, Kyle Rose wrote:<u></u=
><u></u></p>
                              </div>
                              <blockquote style=3D"margin-top:5pt;margin-bo=
ttom:5pt">
                                <p class=3D"MsoNormal">On
                                  Wed, May 27, 2020 at 10:34 AM Ian
                                  Swett &lt;ianswett=3D<a href=3D"mailto:40=
google.com@dmarc.ietf.org" target=3D"_blank">40google.com@dmarc.ietf.org</a=
>&gt;
                                  wrote:<u></u><u></u></p>
                                <div>
                                  <blockquote style=3D"border-color:current=
color currentcolor currentcolor rgb(204,204,204);border-style:none none non=
e solid;border-width:medium medium medium 1pt;padding:0in 0in 0in 6pt;margi=
n:5pt 0in 5pt 4.8pt">
                                    <div>
                                      <p class=3D"MsoNormal">I
                                        was agreeing with MT, but I&#39;m
                                        happy to=C2=A0see some more MUSTs
                                        added if people feel that&#39;d be
                                        helpful.<u></u><u></u></p>
                                    </div>
                                  </blockquote>
                                  <div>
                                    <p class=3D"MsoNormal"><span style=3D"f=
ont-size:12pt">=C2=A0</span><u></u><u></u></p>
                                    <div>
                                      <div>
                                        <p class=3D"MsoNormal"><span style=
=3D"font-size:12pt">Coincidentally,
                                            we were just talking about
                                            this internally at Akamai
                                            yesterday. IMO, an
                                            invariants document isn&#39;t
                                            really helpful if it isn&#39;t
                                            normative, and for it to be
                                            normative it (or a related
                                            practices doc for operators)
                                            really needs to spell out
                                            clear boundaries for
                                            operators with MUSTs..</span><u=
></u><u></u></p>
                                      </div>
                                      <div>
                                        <p class=3D"MsoNormal"><span style=
=3D"font-size:12pt">=C2=A0</span><u></u><u></u></p>
                                      </div>
                                      <div>
                                        <p class=3D"MsoNormal"><span style=
=3D"font-size:12pt">The
                                            example that came up
                                            yesterday was around
                                            operators filtering QUIC in
                                            the event of a DDoS: one
                                            recommendation based on some
                                            conversations going back at
                                            least to Prague 2019 was to
                                            hash packets on 4-tuple and
                                            filter those below a hash
                                            value chosen for a desired
                                            ingress limit instead of
                                            doing what most operators do
                                            with UDP today, which is to
                                            cap UDP throughput and just
                                            drop packets randomly or
                                            tail drop.</span><u></u><u></u>=
</p>
                                      </div>
                                    </div>
                                  </div>
                                </div>
                              </blockquote>
                              <p>Interesting. Did they consider using
                                the CID, or a fraction of it? This looks
                                entirely like the scenario for which we
                                developed stateless retry.<u></u><u></u></p=
>
                              <blockquote style=3D"margin-top:5pt;margin-bo=
ttom:5pt">
                                <div>
                                  <div>
                                    <div>
                                      <div>
                                        <p class=3D"MsoNormal"><span style=
=3D"font-size:12pt">=C2=A0</span><u></u><u></u></p>
                                      </div>
                                      <div>
                                        <p class=3D"MsoNormal"><span style=
=3D"font-size:12pt">This
                                            recommendation certainly
                                            imposes some constraints on
                                            future protocol development
                                            that motivate new
                                            invariants: for instance, it
                                            would preclude sharding a
                                            connection across multiple
                                            source ports (not that there
                                            is necessarily a reason to
                                            do this; it&#39;s just an
                                            example). But more
                                            importantly, it goes beyond
                                            invariants: it&#39;s one among
                                            many practices compatible
                                            with the current set of
                                            invariants, some reasonable
                                            and some terrible.</span><u></u=
><u></u></p>
                                      </div>
                                    </div>
                                  </div>
                                </div>
                              </blockquote>
                              <p>This would break the &quot;preferred
                                address&quot; redirection. Preferred addres=
s
                                migration may or may not be spelled out
                                in the invariants.<u></u><u></u></p>
                              <blockquote style=3D"margin-top:5pt;margin-bo=
ttom:5pt">
                                <div>
                                  <div>
                                    <div>
                                      <div>
                                        <p class=3D"MsoNormal"><span style=
=3D"font-size:12pt">=C2=A0</span><u></u><u></u></p>
                                      </div>
                                      <div>
                                        <p class=3D"MsoNormal"><span style=
=3D"font-size:12pt">Operators
                                            are going to do things to
                                            QUIC traffic, so it would be
                                            good to offer them
                                            recommendations that are
                                            compatible with broad
                                            deployability.</span><u></u><u>=
</u></p>
                                      </div>
                                    </div>
                                  </div>
                                </div>
                              </blockquote>
                              <p>=C2=A0<u></u><u></u></p>
                              <p>Yes, we do need the invariants for
                                that.<u></u><u></u></p>
                              <p>-- Christian Huitema<u></u><u></u></p>
                            </div>
                          </div>
                        </div>
                      </blockquote>
                    </div>
                  </div>
                </div>
              </div>
            </blockquote>
          </div>
        </div>
      </div>
    </blockquote>
  </div>

</blockquote></div></div>

--000000000000edac6605a8771764--


From nobody Fri Jun 19 15:43:06 2020
Return-Path: <huitema@huitema.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 6ABC53A0F07 for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 15:43:04 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.798
X-Spam-Level: 
X-Spam-Status: No, score=-1.798 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, HTTPS_HTTP_MISMATCH=0.1, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=no autolearn_force=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id AkvOR8r0Na7g for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 15:43:01 -0700 (PDT)
Received: from mx36-out10.antispamcloud.com (mx36-out10.antispamcloud.com [209.126.121.30]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 9DF2E3A0F03 for <quic@ietf.org>; Fri, 19 Jun 2020 15:43:01 -0700 (PDT)
Received: from xse22.mail2web.com ([66.113.196.22] helo=xse.mail2web.com) by mx128.antispamcloud.com with esmtp (Exim 4.92) (envelope-from <huitema@huitema.net>) id 1jmPiq-000p5O-Fa for quic@ietf.org; Sat, 20 Jun 2020 00:43:00 +0200
Received: from xsmtp22.mail2web.com (unknown [10.100.68.61]) by xse.mail2web.com (Postfix) with ESMTPS id 49pYjT5Vnrz3TQd for <quic@ietf.org>; Fri, 19 Jun 2020 15:42:53 -0700 (PDT)
Received: from [10.5.2.13] (helo=xmail03.myhosting.com) by xsmtp22.mail2web.com with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:256) (Exim 4.92) (envelope-from <huitema@huitema.net>) id 1jmPin-0006AF-Jq for quic@ietf.org; Fri, 19 Jun 2020 15:42:53 -0700
Received: (qmail 24797 invoked from network); 19 Jun 2020 22:42:53 -0000
Received: from unknown (HELO [192.168.1.107]) (Authenticated-user:_huitema@huitema.net@[172.58.43.64]) (envelope-sender <huitema@huitema.net>) by xmail03.myhosting.com (qmail-ldap-1.03) with ESMTPA for <jared@puck.nether.net>; 19 Jun 2020 22:42:52 -0000
To: Ted Hardie <ted.ietf@gmail.com>
Cc: "Lubashev, Igor" <ilubashe=40akamai.com@dmarc.ietf.org>, Martin Duke <martin.h.duke@gmail.com>, IETF QUIC WG <quic@ietf.org>, Lars Eggert <lars@eggert.org>, Kyle Rose <krose@krose.org>, Ian Swett <ianswett=40google.com@dmarc.ietf.org>, Jared Mauch <jared@puck.nether.net>
References: <CAM4esxQBqfrz24riPQA_VGKcGp_TzW0pqb97KfFMtNdW9pUfDg@mail.gmail.com> <833A693C-62E6-4889-9954-FCE65A839A7C@eggert.org> <CAKcm_gPMO2DtqvKucqVw0zDjSniSOmFD4p1Tp4YLjr9WSWdEUw@mail.gmail.com> <CAJU8_nUN42gGmQof24XD9-EjXedyzcarDyRP8MGe1qW-BZ=+Aw@mail.gmail.com> <9cd91c24-c730-22a4-7aa0-baf61613b3ce@huitema.net> <f4922cdb59014202900de44cc5fea0ff@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQvwkTvpUcu6-+W5zWo22m-R1jvN7DcCpXfuw8Hb55qsw@mail.gmail.com> <95dd02c92b32472d9cab0dd47b98c637@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQxxXn27rZEY75-AsHD5VF0fqiV1VDyeSrzQ=-sM7JNCg@mail.gmail.com> <9c2e300c30f74d1794d11cf4334ea07b@usma1ex-dag1mb5.msg.corp.akamai.com> <2c40f3d9-fa40-9834-ac30-36bc9a1a6303@huitema.net> <CA+9kkMBQt001xOVgT=9G8YOOTOJ+9S=OWDwGEeYuKVm46Cq3iQ@mail.gmail.com>
From: Christian Huitema <huitema@huitema.net>
Autocrypt: addr=huitema@huitema.net; prefer-encrypt=mutual; keydata= mDMEXtavGxYJKwYBBAHaRw8BAQdA1ou9A5MHTP9N3jfsWzlDZ+jPnQkusmc7sfLmWVz1Rmu0 J0NocmlzdGlhbiBIdWl0ZW1hIDxodWl0ZW1hQGh1aXRlbWEubmV0PoiWBBMWCAA+FiEEw3G4 Nwi4QEpAAXUUELAmqKBYtJQFAl7WrxsCGwMFCQlmAYAFCwkIBwIGFQoJCAsCBBYCAwECHgEC F4AACgkQELAmqKBYtJQbMwD/ebj/qnSbthC/5kD5DxZ/Ip0CGJw5QBz/+fJp3R8iAlsBAMjK r2tmyWyJz0CUkVG24WaR5EAJDvgwDv8h22U6QVkAuDgEXtavGxIKKwYBBAGXVQEFAQEHQJoM 6MUAIqpoqdCIiACiEynZf7nlJg2Eu0pXIhbUGONdAwEIB4h+BBgWCAAmFiEEw3G4Nwi4QEpA AXUUELAmqKBYtJQFAl7WrxsCGwwFCQlmAYAACgkQELAmqKBYtJRm2wD7BzeK5gEXSmBcBf0j BYdSaJcXNzx4yPLbP4GnUMAyl2cBAJzcsR4RkwO4dCRqM9CHpVJCwHtbUDJaa55//E0kp+gH
Subject: Re: Is the invariants draft really standards track?
Message-ID: <3bb42dfc-17ba-c5c8-03f7-35428756b4c2@huitema.net>
Date: Fri, 19 Jun 2020 15:42:53 -0700
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:68.0) Gecko/20100101 Thunderbird/68.9.0
MIME-Version: 1.0
In-Reply-To: <CA+9kkMBQt001xOVgT=9G8YOOTOJ+9S=OWDwGEeYuKVm46Cq3iQ@mail.gmail.com>
Content-Type: multipart/alternative; boundary="------------4A390088E034164F5E19A195"
Content-Language: en-US
X-Originating-IP: 66.113.196.22
X-Spampanel-Domain: xsmtpout.mail2web.com
X-Spampanel-Username: 66.113.196.22/32
Authentication-Results: antispamcloud.com; auth=pass smtp.auth=66.113.196.22/32@xsmtpout.mail2web.com
X-Spampanel-Outgoing-Class: unsure
X-Spampanel-Outgoing-Evidence: Combined (0.15)
X-Recommended-Action: accept
X-Filter-ID: Mvzo4OR0dZXEDF/gcnlw0f6LF1GdvkEexklpcFpSF5apSDasLI4SayDByyq9LIhVUZbR67CQ7/vm /hHDJU4RXkTNWdUk1Ol2OGx3IfrIJKywOmJyM1qr8uRnWBrbSAGDcnqpk5VeF3xR4kF6iVwRtbgN zB/4Jkrw1eDLcif59fu2AQXEmZxtT0lc3gN06y8zU7Tmz6iKnkQL9gqsxD347235Nhqq+/HvroPq 8GSPg+60/QPNqXybIny9WGhadIo/+iU2d2HKPwiXQ4BZ4N2yebyme9ldZJ7uNXfg/GfS8fUvP/L5 rCqHDsKZM+xa1iwJX+gRCHfMVnsAk591zk0uilUI+ZL4xWiN8NS6C+dmX6OEdA4u1aThyWrQ/ou2 +v/lmX4Em37yFgrCB6NHRn1g+f3uncIqYSL3lhh5c81YyJqFoLZMmkWsaurVZfvqROaDnDtHb8z5 dpPkEuJ8Snwqla7jUnW3hy14Yji8fo+4xCnSRo4Rcu5Z37rMuDjCny5fE9ykbJ7I9co1MAEE3ruN Xsm8UJsAPvDcVSKtDCYkioPY5Qx4fJOk03R5fJtf/Dv/dkIzS7m4GUpXCY1Y3j3ilUN7TTX3qb0a 8RNcOLCOSd4jWNjicid92+JsmBTBq789v5v+DILSciodxmlf53LGR6hxg06zpR/wTPgbRPrjgV/0 lSfuxANzRU5MAZzTOSGBRPZGlP7x+w/0MFbi1QWNEPKY2AXNZGS5G93aGyH8MqMNONNOB63tZ91H 4Bn0Oix6L+b/Q3An7mTMkg6pGuwvx5xMPnetLBJMh51NiRRoHIAtig2RghsqiEk66dok7aOwmiK7 x42VjdzChZMe6O/DiWiiIzuXMTE3l4bIsk+O50vNMQF6QEd8ADoB3ZqpDw0z08QV3No+S2msRDep v5w/kkG0v17AmegcpQ0tml/sN9lmMy/o83jVXTcfb9k0nLWblJy7uxV6dw8jzlsaNZe6hynMJcjx DydxsJEju76A7X1QIVydqXpZ6MHhiKws9Iiut28r9wo4SqUIg8Yh9hAM0n3LLzx/F2gT3wl8JQJv Bho=
X-Report-Abuse-To: spam@quarantine11.antispamcloud.com
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/Piu7nLZ9CC_lHohZe-7xmyQFkps>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 19 Jun 2020 22:43:05 -0000

This is a multi-part message in MIME format.
--------------4A390088E034164F5E19A195
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable


On 6/19/2020 3:05 PM, Ted Hardie wrote:
> Hi Christian,
>
> A question in-line,
>
> On Fri, Jun 19, 2020 at 2:49 PM Christian Huitema <huitema@huitema.net
> <mailto:huitema@huitema.net>> wrote:
>
>     When under DOS attack, you want to "minimize blowback", i.e., as
>     much as possible avoid generating packets in response to attack
>     traffic. So, yes, a server may choose to not send stateless resets
>     to anyone when under attack; in fact, my recommendation would be
>     that a server SHOULD NOT send stateless resets to anyone when
>     under attack.
>
>     That said, Igor raised an interesting point about return traffic.
>     It would be very nice if DOS protection boxes could distinguish
>     between "validated traffic" that the server presumably intends to
>     process, and "unsolicited traffic" that will just consume
>     resource. The box can then reserve some share of the resource for
>     validated traffic, and place the rest of the traffic in a lower
>     priority queue. Fine, but there needs to be a test. The classic
>     test is that incoming traffic is "validated" if the protection box
>     can match it with return traffic coming from the server -- for
>     some definition of matching.
>
>     From that point of view, stateless reset is definitely not
>     helpful. But problematic traffic goes beyond that. The server will
>     reply to a client's initial with a server's initial packet. Does
>     that validate the response traffic? OK, maybe the protection box
>     can programmed to only validate traffic if it sees 1RTT packets.
>     But many servers will send 1-RTT packets as 0.5 RTT. Does that
>     validate the response traffic?
>
>     We might say that traffic is validated when the handshake is
>     confirmed, but the protection box does not understand the TLS
>     handshake, it just sees packet types and packet sizes. It cannot
>     distinguish between 0.5RTT data and 1RTT data, and thus the
>     closest approximation of "validation" would be seeing more than an
>     initial window worth of traffic coming back from the server. That
>     does not sound great.
>
>     On the other hand, things get much better if the server under
>     attack can adopt a defensive posture and help the DOS protection
>     box do its job. Suppose that a server can detect that it is under
>     attack -- or be explicitly configured so. The simplest defensive
>     posture would be to (1) disable stateless reset and (2) not send
>     any 0.5RTT packet, including in response to 0-RTT. The protection
>     boxes can at that point take the 1-RTT packets from the server as
>     indicating validation.
>
> Perhaps I'm misreading you here, but this sounds like you expect the
> protection boxes to be able to distinguish when servers see themselves
> as under DOS from when they don't, so that they can tell that the lack
> of 0.5RTT is an indication of an attack response.=C2=A0 Given distribut=
ion
> patterns of DOS attacks, I'm struggling to see whether that will
> commonly be the case.

Maybe I wasn't too clear. I don't think that DOS protection boxes are
trying to take directions from servers. I also don't believe that DOS
protection boxes have any practical means of distinguishing between 0.5
RTT and 1 RTT traffic. That might actually be a problem.

>
> You could, of course, always put handshake traffic into low-priority
> queues until you see the 1-RTT packets that validate the server's
> interest.=C2=A0 That would make 1-RTT traffic effectively a path signal=
 of
> validation.=C2=A0

Yes, that's more or less what I was considering. But of course it only
works if the server refrains from sending 1RTT packets until the
handshake is confirmed. Otherwise, the protection boxes will have to
count the number of 1RTT packets coming back from the server, or maybe
the amount of data coming from the server, and only consider the traffic
validated if number of packets and amount of data are larger than common
values of IW. You might end up with 3 classes instead of 2 -- validated
if 1RTT > IW, validation in progress if 1RTT see but < IW, not validated
yet if no 1RTT seen.

> My concern with that is that using those low priority queues during
> the handshake phase seems likely to result in worse latency and
> increased risk of packet loss (which can be tricky during that
> phase).=C2=A0 That seems a heavy price to pay during non-attack times f=
or
> better protection when under attack.

The problem of the DOS protection box is to drop as much traffic as
possible while still letting some good in, and also without causing
performance hits when there are no attacks. There are indeed some thorny
traffic management issues there. More information helps, and Ben's
suggestion to look at DOTS is certainly on point.

I am just doing a thought experiment so far. One immediate observation
is that in the absence of other data, DOS protection designers are
likely to look at packet types, and certainly reason about 1RTT packets
versus long header packets. This is definitely an ossification risk.

The other observation is that we have to be careful about the realism of
thought experiments. If I remember correctly, the payload of the Blaster
worm was something like:

=C2=A0=C2=A0=C2=A0 On each bot

=C2=A0=C2=A0=C2=A0 =C2=A0=C2=A0=C2=A0 Open 256 threads

=C2=A0=C2=A0=C2=A0 =C2=A0=C2=A0=C2=A0 =C2=A0=C2=A0=C2=A0 In each thread, =
loop on "GET some large page from the server"

Reasoning about validated traffic would not stop that. But reasoning
about validated traffic forces the attacker to disclose the "real" IP
addresses of the attacking bots, which then enables a second line of
defense.

-- Christian Huitema



>
> Am I misunderstanding how this is distinguished?
>
> Clue appreciated,
>
> Ted
>
> =C2=A0
>
>     Maybe we should specify that.
>
>     -- Christian Huitema
>
>     On 6/19/2020 1:42 PM, Lubashev, Igor wrote:
>>
>>     > There is no need for servers to decrypt CIDs in QUIC-LB.
>>     Presumably the server has a lookup table for its CIDs.
>>
>>     =C2=A0
>>
>>     Sending a stateless reset in response to a junk packet would cost
>>     more CPU than verifying CID integrity.=C2=A0 But, yes, a server ma=
y
>>     choose to not send stateless resets to anyone when under attack.
>>
>>     =C2=A0
>>
>>     =C2=A0
>>
>>     *From:* Martin Duke <martin.h.duke@gmail.com>
>>     <mailto:martin.h.duke@gmail.com>
>>     *Sent:* Friday, June 19, 2020 2:44 PM
>>
>>     >Unfortunately, Retry system protects only server's memory state
>>     and some CPU cycles spent on crypto.=C2=A0 (Servers still need to
>>     decrypt CID to decide it is invalid, and if the attacker is
>>     clever enough to establish one valid connection and use that CID
>>     in a flood, the server will also be decrypting packets.)=C2=A0=C2=A0=

>>
>>     =C2=A0
>>
>>     There is no need for servers to decrypt CIDs in QUIC-LB.
>>     Presumably the server has a lookup table for its CIDs.
>>
>>     =C2=A0
>>
>>     It is true that Retry Services (and indeed, the Retry concept as
>>     a whole) does nothing to protect network capacity.
>>
>>     =C2=A0
>>
>>     On Fri, Jun 19, 2020 at 8:08 AM Lubashev, Igor
>>     <ilubashe@akamai.com <mailto:ilubashe@akamai.com>> wrote:
>>
>>         It looks like
>>         https://tools.ietf.org/html/draft-ietf-quic-load-balancers-02#=
section-5
>>         <https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__tools.=
ietf.org_html_draft-2Dietf-2Dquic-2Dload-2Dbalancers-2D02-23section-2D5&d=
=3DDwMFaQ&c=3D96ZbZZcaMF4w0F4jpN6LZg&r=3DDjn3bQ5uNJDPM_2skfL3rW1tzcIxyjUZ=
dn_m55KPmlo&m=3DeTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTKljvj-ys&s=3D1x7AN2a51=
X_zV5xSyK0uCL8vZ5cugD3n4lbFWZDqVQg&e=3D>
>>         is an excellent discussion of Retry mechanics.=C2=A0 It defini=
tely
>>         deserves a reference from this manageability draft.
>>
>>         =C2=A0
>>
>>         The Retry mechanisms described in LB draft are all
>>         cooperating boxes, and servers must be aware of them.=C2=A0
>>         Unfortunately, Retry system protects only server's memory
>>         state and some CPU cycles spent on crypto.=C2=A0 (Servers stil=
l
>>         need to decrypt CID to decide it is invalid, and if the
>>         attacker is clever enough to establish one valid connection
>>         and use that CID in a flood, the server will also be
>>         decrypting packets.)=C2=A0 Retry does nothing to protect netwo=
rk
>>         resources.
>>
>>         =C2=A0
>>
>>         The PR I opened (https://github.com/quicwg/ops-drafts/pull/94
>>         <https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__github=
=2Ecom_quicwg_ops-2Ddrafts_pull_94&d=3DDwMFaQ&c=3D96ZbZZcaMF4w0F4jpN6LZg&=
r=3DDjn3bQ5uNJDPM_2skfL3rW1tzcIxyjUZdn_m55KPmlo&m=3DeTT-BoZ1fMitywKRVSxpU=
3js0lhO0qkspYTKljvj-ys&s=3DVpvhQ9n79rANIk3O5Sm7PZyToFQEennoPt9iqFpWbq8&e=3D=
>)
>>         is about uncooperating devices that try to mitigate
>>         volumetric network attacks.
>>
>>         =C2=A0
>>
>>         =C2=A0
>>
>>         *From:* Martin Duke <martin.h.duke@gmail.com
>>         <mailto:martin.h.duke@gmail.com>>
>>         *Sent:* Wednesday, June 17, 2020 8:16 PM
>>
>>         Hi Igor, you might want to check out the "Retry Services" bit
>>         of the QUIC-LB draft. This has something to do with the DDoS
>>         use case you discuss.
>>
>>         =C2=A0
>>
>>         On Wed, May 27, 2020 at 9:07 AM Lubashev, Igor
>>         <ilubashe@akamai.com <mailto:ilubashe@akamai.com>> wrote:
>>
>>             I=E2=80=99m working on a manageability draft PR for this (=
how to
>>             rate limit UDP to reduce disruption to QUIC if you have
>>             to rate limit UDP).=C2=A0 ETA end of the week (if I do not=
 get
>>             pulled into something again).
>>
>>             =C2=A0
>>
>>             The relevant observation is that DDoS with UDP that is
>>             indistinguishable from QUIC will happen.=C2=A0 UDP is alre=
ady
>>             the most prevalent DDoS vector, since it is easy for a
>>             compromised non-admin app to send a flood of huge UDP
>>             packets (with TCP you get throttled by the congestion
>>             controller).=C2=A0 So there WILL be DDoS protection device=
s
>>             out there to try to mitigate the problem, possibly by
>>             observing both directions of the flow and deciding
>>             whether a packet belongs to a valid flow or not.
>>
>>             =C2=A0
>>
>>             Since such middle boxes will be created, the more
>>             explicit and normative Invariants are about what one can
>>             expect, the less such middle boxes may decide for
>>             themselves.=C2=A0 For example (I did not think long about =
it),
>>             if some elements of path validation could land into
>>             Invariants (roughly, =E2=80=9Cno more than X packets/bytes=
 can be
>>             sent on a new path w/o a return packet=E2=80=9D), a DDoS m=
iddle
>>             box may use this fact and active connection migration
>>             might still have a chance during an attack (NAT rebinding
>>             could be linked by DDoS boxes to an old connection via
>>             unchanged CID).
>>
>>             =C2=A0
>>
>>               * Igor
>>
>>             =C2=A0
>>
>>             =C2=A0
>>
>>             *From:* Christian Huitema <huitema@huitema.net
>>             <mailto:huitema@huitema.net>>
>>             *Sent:* Wednesday, May 27, 2020 11:34 AM
>>
>>             On 5/27/2020 8:28 AM, Kyle Rose wrote:
>>
>>                 On Wed, May 27, 2020 at 10:34 AM Ian Swett
>>                 <ianswett=3D40google.com@dmarc.ietf.org
>>                 <mailto:40google.com@dmarc.ietf.org>> wrote:
>>
>>                     I was agreeing with MT, but I'm happy to=C2=A0see =
some
>>                     more MUSTs added if people feel that'd be helpful.=

>>
>>                 =C2=A0
>>
>>                 Coincidentally, we were just talking about this
>>                 internally at Akamai yesterday. IMO, an invariants
>>                 document isn't really helpful if it isn't normative,
>>                 and for it to be normative it (or a related practices
>>                 doc for operators) really needs to spell out clear
>>                 boundaries for operators with MUSTs..
>>
>>                 =C2=A0
>>
>>                 The example that came up yesterday was around
>>                 operators filtering QUIC in the event of a DDoS: one
>>                 recommendation based on some conversations going back
>>                 at least to Prague 2019 was to hash packets on
>>                 4-tuple and filter those below a hash value chosen
>>                 for a desired ingress limit instead of doing what
>>                 most operators do with UDP today, which is to cap UDP
>>                 throughput and just drop packets randomly or tail drop=
=2E
>>
>>             Interesting. Did they consider using the CID, or a
>>             fraction of it? This looks entirely like the scenario for
>>             which we developed stateless retry.
>>
>>                 =C2=A0
>>
>>                 This recommendation certainly imposes some
>>                 constraints on future protocol development that
>>                 motivate new invariants: for instance, it would
>>                 preclude sharding a connection across multiple source
>>                 ports (not that there is necessarily a reason to do
>>                 this; it's just an example). But more importantly, it
>>                 goes beyond invariants: it's one among many practices
>>                 compatible with the current set of invariants, some
>>                 reasonable and some terrible.
>>
>>             This would break the "preferred address" redirection.
>>             Preferred address migration may or may not be spelled out
>>             in the invariants.
>>
>>                 =C2=A0
>>
>>                 Operators are going to do things to QUIC traffic, so
>>                 it would be good to offer them recommendations that
>>                 are compatible with broad deployability.
>>
>>             =C2=A0
>>
>>             Yes, we do need the invariants for that.
>>
>>             -- Christian Huitema
>>

--------------4A390088E034164F5E19A195
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: 8bit

<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <p><br>
    </p>
    <div class="moz-cite-prefix">On 6/19/2020 3:05 PM, Ted Hardie wrote:<br>
    </div>
    <blockquote type="cite"
cite="mid:CA+9kkMBQt001xOVgT=9G8YOOTOJ+9S=OWDwGEeYuKVm46Cq3iQ@mail.gmail.com">
      <meta http-equiv="content-type" content="text/html; charset=UTF-8">
      <div dir="ltr">
        <div dir="ltr">
          <div>Hi Christian,</div>
          <div><br>
          </div>
          <div>A question in-line,<br>
          </div>
        </div>
        <br>
        <div class="gmail_quote">
          <div dir="ltr" class="gmail_attr">On Fri, Jun 19, 2020 at 2:49
            PM Christian Huitema &lt;<a
              href="mailto:huitema@huitema.net" moz-do-not-send="true">huitema@huitema.net</a>&gt;
            wrote:<br>
          </div>
          <blockquote class="gmail_quote" style="margin:0px 0px 0px
            0.8ex;border-left:1px solid
            rgb(204,204,204);padding-left:1ex">
            <div>
              <p>When under DOS attack, you want to "minimize blowback",
                i.e., as much as possible avoid generating packets in
                response to attack traffic. So, yes, a server may choose
                to not send stateless resets to anyone when under
                attack; in fact, my recommendation would be that a
                server SHOULD NOT send stateless resets to anyone when
                under attack.</p>
              <p>That said, Igor raised an interesting point about
                return traffic. It would be very nice if DOS protection
                boxes could distinguish between "validated traffic" that
                the server presumably intends to process, and
                "unsolicited traffic" that will just consume resource.
                The box can then reserve some share of the resource for
                validated traffic, and place the rest of the traffic in
                a lower priority queue. Fine, but there needs to be a
                test. The classic test is that incoming traffic is
                "validated" if the protection box can match it with
                return traffic coming from the server -- for some
                definition of matching.</p>
              <p>From that point of view, stateless reset is definitely
                not helpful. But problematic traffic goes beyond that.
                The server will reply to a client's initial with a
                server's initial packet. Does that validate the response
                traffic? OK, maybe the protection box can programmed to
                only validate traffic if it sees 1RTT packets. But many
                servers will send 1-RTT packets as 0.5 RTT. Does that
                validate the response traffic?</p>
              <p>We might say that traffic is validated when the
                handshake is confirmed, but the protection box does not
                understand the TLS handshake, it just sees packet types
                and packet sizes. It cannot distinguish between 0.5RTT
                data and 1RTT data, and thus the closest approximation
                of "validation" would be seeing more than an initial
                window worth of traffic coming back from the server.
                That does not sound great.</p>
              <p>On the other hand, things get much better if the server
                under attack can adopt a defensive posture and help the
                DOS protection box do its job. Suppose that a server can
                detect that it is under attack -- or be explicitly
                configured so. The simplest defensive posture would be
                to (1) disable stateless reset and (2) not send any
                0.5RTT packet, including in response to 0-RTT. The
                protection boxes can at that point take the 1-RTT
                packets from the server as indicating validation.</p>
            </div>
          </blockquote>
          <div>Perhaps I'm misreading you here, but this sounds like you
            expect the protection boxes to be able to distinguish when
            servers see themselves as under DOS from when they don't, so
            that they can tell that the lack of 0.5RTT is an indication
            of an attack response.  Given distribution patterns of DOS
            attacks, I'm struggling to see whether that will commonly be
            the case.</div>
        </div>
      </div>
    </blockquote>
    <p>Maybe I wasn't too clear. I don't think that DOS protection boxes
      are trying to take directions from servers. I also don't believe
      that DOS protection boxes have any practical means of
      distinguishing between 0.5 RTT and 1 RTT traffic. That might
      actually be a problem.<br>
    </p>
    <blockquote type="cite"
cite="mid:CA+9kkMBQt001xOVgT=9G8YOOTOJ+9S=OWDwGEeYuKVm46Cq3iQ@mail.gmail.com">
      <div dir="ltr">
        <div class="gmail_quote">
          <div><br>
          </div>
          <div>You could, of course, always put handshake traffic into
            low-priority queues until you see the 1-RTT packets that
            validate the server's interest.  That would make 1-RTT
            traffic effectively a path signal of validation.  </div>
        </div>
      </div>
    </blockquote>
    <p>Yes, that's more or less what I was considering. But of course it
      only works if the server refrains from sending 1RTT packets until
      the handshake is confirmed. Otherwise, the protection boxes will
      have to count the number of 1RTT packets coming back from the
      server, or maybe the amount of data coming from the server, and
      only consider the traffic validated if number of packets and
      amount of data are larger than common values of IW. You might end
      up with 3 classes instead of 2 -- validated if 1RTT &gt; IW,
      validation in progress if 1RTT see but &lt; IW, not validated yet
      if no 1RTT seen.<br>
    </p>
    <blockquote type="cite"
cite="mid:CA+9kkMBQt001xOVgT=9G8YOOTOJ+9S=OWDwGEeYuKVm46Cq3iQ@mail.gmail.com">
      <div dir="ltr">
        <div class="gmail_quote">
          <div>My concern with that is that using those low priority
            queues during the handshake phase seems likely to result in
            worse latency and increased risk of packet loss (which can
            be tricky during that phase).  That seems a heavy price to
            pay during non-attack times for better protection when under
            attack.</div>
        </div>
      </div>
    </blockquote>
    <p>The problem of the DOS protection box is to drop as much traffic
      as possible while still letting some good in, and also without
      causing performance hits when there are no attacks. There are
      indeed some thorny traffic management issues there. More
      information helps, and Ben's suggestion to look at DOTS is
      certainly on point.</p>
    <p>I am just doing a thought experiment so far. One immediate
      observation is that in the absence of other data, DOS protection
      designers are likely to look at packet types, and certainly reason
      about 1RTT packets versus long header packets. This is definitely
      an ossification risk.</p>
    <p>The other observation is that we have to be careful about the
      realism of thought experiments. If I remember correctly, the
      payload of the Blaster worm was something like:</p>
    <p>    On each bot</p>
    <p>        Open 256 threads</p>
    <p>            In each thread, loop on "GET some large page from the
      server"</p>
    <p>Reasoning about validated traffic would not stop that. But
      reasoning about validated traffic forces the attacker to disclose
      the "real" IP addresses of the attacking bots, which then enables
      a second line of defense.<br>
    </p>
    <p>-- Christian Huitema<br>
    </p>
    <p><br>
    </p>
    <p><br>
    </p>
    <blockquote type="cite"
cite="mid:CA+9kkMBQt001xOVgT=9G8YOOTOJ+9S=OWDwGEeYuKVm46Cq3iQ@mail.gmail.com">
      <div dir="ltr">
        <div class="gmail_quote">
          <div><br>
          </div>
          <div>Am I misunderstanding how this is distinguished?</div>
          <div><br>
          </div>
          <div>Clue appreciated,</div>
          <div><br>
          </div>
          <div>Ted<br>
          </div>
          <div><br>
          </div>
          <div> </div>
          <blockquote class="gmail_quote" style="margin:0px 0px 0px
            0.8ex;border-left:1px solid
            rgb(204,204,204);padding-left:1ex">
            <div>
              <p>Maybe we should specify that.</p>
              <p>-- Christian Huitema<br>
              </p>
              <div>On 6/19/2020 1:42 PM, Lubashev, Igor wrote:<br>
              </div>
              <blockquote type="cite">
                <div>
                  <p class="MsoNormal">&gt; There is no need for servers
                    to decrypt CIDs in QUIC-LB. Presumably the server
                    has a lookup table for its CIDs.</p>
                  <p class="MsoNormal"> </p>
                  <p class="MsoNormal">Sending a stateless reset in
                    response to a junk packet would cost more CPU than
                    verifying CID integrity.  But, yes, a server may
                    choose to not send stateless resets to anyone when
                    under attack.</p>
                  <p class="MsoNormal"> </p>
                  <p class="MsoNormal"> </p>
                  <div style="border-color:currentcolor currentcolor
                    currentcolor blue;border-style:none none none
                    solid;border-width:medium medium medium
                    1.5pt;padding:0in 0in 0in 4pt">
                    <div>
                      <div style="border-color:rgb(225,225,225)
                        currentcolor currentcolor;border-style:solid
                        none none;border-width:1pt medium
                        medium;padding:3pt 0in 0in">
                        <p class="MsoNormal"><b>From:</b> Martin Duke <a
                            href="mailto:martin.h.duke@gmail.com"
                            target="_blank" moz-do-not-send="true">&lt;martin.h.duke@gmail.com&gt;</a>
                          <br>
                          <b>Sent:</b> Friday, June 19, 2020 2:44 PM<br>
                          <br>
                        </p>
                      </div>
                    </div>
                    <div>
                      <p class="MsoNormal">&gt;Unfortunately, Retry
                        system protects only server's memory state and
                        some CPU cycles spent on crypto.  (Servers still
                        need to decrypt CID to decide it is invalid, and
                        if the attacker is clever enough to establish
                        one valid connection and use that CID in a
                        flood, the server will also be decrypting
                        packets.)  </p>
                      <div>
                        <p class="MsoNormal"> </p>
                      </div>
                      <div>
                        <p class="MsoNormal">There is no need for
                          servers to decrypt CIDs in QUIC-LB. Presumably
                          the server has a lookup table for its CIDs.</p>
                      </div>
                      <div>
                        <p class="MsoNormal"> </p>
                      </div>
                      <div>
                        <p class="MsoNormal">It is true that Retry
                          Services (and indeed, the Retry concept as a
                          whole) does nothing to protect network
                          capacity.</p>
                      </div>
                    </div>
                    <p class="MsoNormal"> </p>
                    <div>
                      <div>
                        <p class="MsoNormal">On Fri, Jun 19, 2020 at
                          8:08 AM Lubashev, Igor &lt;<a
                            href="mailto:ilubashe@akamai.com"
                            target="_blank" moz-do-not-send="true">ilubashe@akamai.com</a>&gt;
                          wrote:</p>
                      </div>
                      <blockquote style="border-color:currentcolor
                        currentcolor currentcolor
                        rgb(204,204,204);border-style:none none none
                        solid;border-width:medium medium medium
                        1pt;padding:0in 0in 0in
                        6pt;margin-left:4.8pt;margin-right:0in">
                        <div>
                          <div>
                            <p class="MsoNormal">It looks like <a
href="https://urldefense.proofpoint.com/v2/url?u=https-3A__tools.ietf.org_html_draft-2Dietf-2Dquic-2Dload-2Dbalancers-2D02-23section-2D5&amp;d=DwMFaQ&amp;c=96ZbZZcaMF4w0F4jpN6LZg&amp;r=Djn3bQ5uNJDPM_2skfL3rW1tzcIxyjUZdn_m55KPmlo&amp;m=eTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTKljvj-ys&amp;s=1x7AN2a51X_zV5xSyK0uCL8vZ5cugD3n4lbFWZDqVQg&amp;e="
                                target="_blank" moz-do-not-send="true">
https://tools.ietf.org/html/draft-ietf-quic-load-balancers-02#section-5</a>
                              is an excellent discussion of Retry
                              mechanics.  It definitely deserves a
                              reference from this manageability draft.</p>
                            <p class="MsoNormal"> </p>
                            <p class="MsoNormal">The Retry mechanisms
                              described in LB draft are all cooperating
                              boxes, and servers must be aware of them. 
                              Unfortunately, Retry system protects only
                              server's memory state and some CPU cycles
                              spent on crypto.  (Servers still need to
                              decrypt CID to decide it is invalid, and
                              if the attacker is clever enough to
                              establish one valid connection and use
                              that CID in a flood, the server will also
                              be decrypting packets.)  Retry does
                              nothing to protect network resources.</p>
                            <p class="MsoNormal"> </p>
                            <p class="MsoNormal">The PR I opened (<a
href="https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_quicwg_ops-2Ddrafts_pull_94&amp;d=DwMFaQ&amp;c=96ZbZZcaMF4w0F4jpN6LZg&amp;r=Djn3bQ5uNJDPM_2skfL3rW1tzcIxyjUZdn_m55KPmlo&amp;m=eTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTKljvj-ys&amp;s=VpvhQ9n79rANIk3O5Sm7PZyToFQEennoPt9iqFpWbq8&amp;e="
                                target="_blank" moz-do-not-send="true">https://github.com/quicwg/ops-drafts/pull/94</a>)
                              is about uncooperating devices that try to
                              mitigate volumetric network attacks.</p>
                            <p class="MsoNormal"> </p>
                            <p class="MsoNormal"> </p>
                            <div style="border-color:currentcolor
                              currentcolor currentcolor
                              blue;border-style:none none none
                              solid;border-width:medium medium medium
                              1.5pt;padding:0in 0in 0in 4pt">
                              <div>
                                <div
                                  style="border-color:rgb(225,225,225)
                                  currentcolor
                                  currentcolor;border-style:solid none
                                  none;border-width:1pt medium
                                  medium;padding:3pt 0in 0in">
                                  <p class="MsoNormal"
                                    style="margin-bottom:12pt"><b>From:</b>
                                    Martin Duke &lt;<a
                                      href="mailto:martin.h.duke@gmail.com"
                                      target="_blank"
                                      moz-do-not-send="true">martin.h.duke@gmail.com</a>&gt;
                                    <br>
                                    <b>Sent:</b> Wednesday, June 17,
                                    2020 8:16 PM</p>
                                </div>
                              </div>
                              <div>
                                <p class="MsoNormal">Hi Igor, you might
                                  want to check out the "Retry Services"
                                  bit of the QUIC-LB draft. This has
                                  something to do with the DDoS use case
                                  you discuss.</p>
                              </div>
                              <p class="MsoNormal"> </p>
                              <div>
                                <div>
                                  <p class="MsoNormal">On Wed, May 27,
                                    2020 at 9:07 AM Lubashev, Igor &lt;<a
                                      href="mailto:ilubashe@akamai.com"
                                      target="_blank"
                                      moz-do-not-send="true">ilubashe@akamai.com</a>&gt;
                                    wrote:</p>
                                </div>
                                <blockquote
                                  style="border-color:currentcolor
                                  currentcolor currentcolor
                                  rgb(204,204,204);border-style:none
                                  none none solid;border-width:medium
                                  medium medium 1pt;padding:0in 0in 0in
                                  6pt;margin:5pt 0in 5pt 4.8pt">
                                  <div>
                                    <div>
                                      <p class="MsoNormal">I’m working
                                        on a manageability draft PR for
                                        this (how to rate limit UDP to
                                        reduce disruption to QUIC if you
                                        have to rate limit UDP).  ETA
                                        end of the week (if I do not get
                                        pulled into something again).</p>
                                      <p class="MsoNormal"> </p>
                                      <p class="MsoNormal">The relevant
                                        observation is that DDoS with
                                        UDP that is indistinguishable
                                        from QUIC will happen.  UDP is
                                        already the most prevalent DDoS
                                        vector, since it is easy for a
                                        compromised non-admin app to
                                        send a flood of huge UDP packets
                                        (with TCP you get throttled by
                                        the congestion controller).  So
                                        there WILL be DDoS protection
                                        devices out there to try to
                                        mitigate the problem, possibly
                                        by observing both directions of
                                        the flow and deciding whether a
                                        packet belongs to a valid flow
                                        or not.</p>
                                      <p class="MsoNormal"> </p>
                                      <p class="MsoNormal">Since such
                                        middle boxes will be created,
                                        the more explicit and normative
                                        Invariants are about what one
                                        can expect, the less such middle
                                        boxes may decide for
                                        themselves.  For example (I did
                                        not think long about it), if
                                        some elements of path validation
                                        could land into Invariants
                                        (roughly, “no more than X
                                        packets/bytes can be sent on a
                                        new path w/o a return packet”),
                                        a DDoS middle box may use this
                                        fact and active connection
                                        migration might still have a
                                        chance during an attack (NAT
                                        rebinding could be linked by
                                        DDoS boxes to an old connection
                                        via unchanged CID).</p>
                                      <p class="MsoNormal"> </p>
                                      <ul type="disc">
                                        <li class="MsoNormal"> Igor</li>
                                      </ul>
                                      <p class="MsoNormal"> </p>
                                      <p class="MsoNormal"> </p>
                                      <div
                                        style="border-color:currentcolor
                                        currentcolor currentcolor
                                        blue;border-style:none none none
                                        solid;border-width:medium medium
                                        medium 1.5pt;padding:0in 0in 0in
                                        4pt">
                                        <p class="MsoNormal"
                                          style="margin-bottom:12pt"><b>From:</b>
                                          Christian Huitema &lt;<a
                                            href="mailto:huitema@huitema.net"
                                            target="_blank"
                                            moz-do-not-send="true">huitema@huitema.net</a>&gt;
                                          <br>
                                          <b>Sent:</b> Wednesday, May
                                          27, 2020 11:34 AM</p>
                                        <div>
                                          <p class="MsoNormal">On
                                            5/27/2020 8:28 AM, Kyle Rose
                                            wrote:</p>
                                        </div>
                                        <blockquote
                                          style="margin-top:5pt;margin-bottom:5pt">
                                          <p class="MsoNormal">On Wed,
                                            May 27, 2020 at 10:34 AM Ian
                                            Swett &lt;ianswett=<a
                                              href="mailto:40google.com@dmarc.ietf.org"
                                              target="_blank"
                                              moz-do-not-send="true">40google.com@dmarc.ietf.org</a>&gt;
                                            wrote:</p>
                                          <div>
                                            <blockquote
                                              style="border-color:currentcolor
                                              currentcolor currentcolor
rgb(204,204,204);border-style:none none none solid;border-width:medium
                                              medium medium
                                              1pt;padding:0in 0in 0in
                                              6pt;margin:5pt 0in 5pt
                                              4.8pt">
                                              <div>
                                                <p class="MsoNormal">I
                                                  was agreeing with MT,
                                                  but I'm happy to see
                                                  some more MUSTs added
                                                  if people feel that'd
                                                  be helpful.</p>
                                              </div>
                                            </blockquote>
                                            <div>
                                              <p class="MsoNormal"><span
                                                  style="font-size:12pt"> </span></p>
                                              <div>
                                                <div>
                                                  <p class="MsoNormal"><span
style="font-size:12pt">Coincidentally, we were just talking about this
                                                      internally at
                                                      Akamai yesterday.
                                                      IMO, an invariants
                                                      document isn't
                                                      really helpful if
                                                      it isn't
                                                      normative, and for
                                                      it to be normative
                                                      it (or a related
                                                      practices doc for
                                                      operators) really
                                                      needs to spell out
                                                      clear boundaries
                                                      for operators with
                                                      MUSTs..</span></p>
                                                </div>
                                                <div>
                                                  <p class="MsoNormal"><span
style="font-size:12pt"> </span></p>
                                                </div>
                                                <div>
                                                  <p class="MsoNormal"><span
style="font-size:12pt">The example that came up yesterday was around
                                                      operators
                                                      filtering QUIC in
                                                      the event of a
                                                      DDoS: one
                                                      recommendation
                                                      based on some
                                                      conversations
                                                      going back at
                                                      least to Prague
                                                      2019 was to hash
                                                      packets on 4-tuple
                                                      and filter those
                                                      below a hash value
                                                      chosen for a
                                                      desired ingress
                                                      limit instead of
                                                      doing what most
                                                      operators do with
                                                      UDP today, which
                                                      is to cap UDP
                                                      throughput and
                                                      just drop packets
                                                      randomly or tail
                                                      drop.</span></p>
                                                </div>
                                              </div>
                                            </div>
                                          </div>
                                        </blockquote>
                                        <p>Interesting. Did they
                                          consider using the CID, or a
                                          fraction of it? This looks
                                          entirely like the scenario for
                                          which we developed stateless
                                          retry.</p>
                                        <blockquote
                                          style="margin-top:5pt;margin-bottom:5pt">
                                          <div>
                                            <div>
                                              <div>
                                                <div>
                                                  <p class="MsoNormal"><span
style="font-size:12pt"> </span></p>
                                                </div>
                                                <div>
                                                  <p class="MsoNormal"><span
style="font-size:12pt">This recommendation certainly imposes some
                                                      constraints on
                                                      future protocol
                                                      development that
                                                      motivate new
                                                      invariants: for
                                                      instance, it would
                                                      preclude sharding
                                                      a connection
                                                      across multiple
                                                      source ports (not
                                                      that there is
                                                      necessarily a
                                                      reason to do this;
                                                      it's just an
                                                      example). But more
                                                      importantly, it
                                                      goes beyond
                                                      invariants: it's
                                                      one among many
                                                      practices
                                                      compatible with
                                                      the current set of
                                                      invariants, some
                                                      reasonable and
                                                      some terrible.</span></p>
                                                </div>
                                              </div>
                                            </div>
                                          </div>
                                        </blockquote>
                                        <p>This would break the
                                          "preferred address"
                                          redirection. Preferred address
                                          migration may or may not be
                                          spelled out in the invariants.</p>
                                        <blockquote
                                          style="margin-top:5pt;margin-bottom:5pt">
                                          <div>
                                            <div>
                                              <div>
                                                <div>
                                                  <p class="MsoNormal"><span
style="font-size:12pt"> </span></p>
                                                </div>
                                                <div>
                                                  <p class="MsoNormal"><span
style="font-size:12pt">Operators are going to do things to QUIC traffic,
                                                      so it would be
                                                      good to offer them
                                                      recommendations
                                                      that are
                                                      compatible with
                                                      broad
                                                      deployability.</span></p>
                                                </div>
                                              </div>
                                            </div>
                                          </div>
                                        </blockquote>
                                        <p> </p>
                                        <p>Yes, we do need the
                                          invariants for that.</p>
                                        <p>-- Christian Huitema</p>
                                      </div>
                                    </div>
                                  </div>
                                </blockquote>
                              </div>
                            </div>
                          </div>
                        </div>
                      </blockquote>
                    </div>
                  </div>
                </div>
              </blockquote>
            </div>
          </blockquote>
        </div>
      </div>
    </blockquote>
  </body>
</html>

--------------4A390088E034164F5E19A195--


From nobody Fri Jun 19 16:00:32 2020
Return-Path: <martin.h.duke@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id CE49D3A0F21 for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 16:00:30 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.997
X-Spam-Level: 
X-Spam-Status: No, score=-1.997 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, HTTPS_HTTP_MISMATCH=0.1, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=unavailable autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Z-xpPRPtiwy9 for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 16:00:26 -0700 (PDT)
Received: from mail-io1-xd2d.google.com (mail-io1-xd2d.google.com [IPv6:2607:f8b0:4864:20::d2d]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 74A313A0F1A for <quic@ietf.org>; Fri, 19 Jun 2020 16:00:26 -0700 (PDT)
Received: by mail-io1-xd2d.google.com with SMTP id s18so13242880ioe.2 for <quic@ietf.org>; Fri, 19 Jun 2020 16:00:26 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=EPjKCH+7rnwVH/tUTHzwtraZ7RI3+w9OizRfBZ60n6I=; b=S3gKX7uY6o4Vm0Q1FwKD8qgCZTYTlflBiyHbUurUrkhBie2BP881HsiQ6ItPPACw6f ByklO8o6QCrM3W6USPFer9PYxK80+JaDirbqB9dYTOKsPLEWcpIa74YiY+qSZCObsCGs CyzWqryOmqRhRj16ANNr9HvRqIkNLxlL05J+U0J2jALluxea5F3thXGIyt5wx1QImSvB ztIc7RlCXr7PM/Cf/Z+X155CJHvXyp32vJ6PDPnjOafk11WwGEC1lepNV3r3694MmLPP loSpBJG0+4y74bUmpoe0PYmTa5yCcocXI7PcXvyEP/r6c/bVYumk876R+Z5B03T+12R3 L1fQ==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=EPjKCH+7rnwVH/tUTHzwtraZ7RI3+w9OizRfBZ60n6I=; b=KAt0EXMBdJNtviOZeJQS/qDcS3dZpqhG74fpAaddPriWsJmV++BxzzypUVDsmOfgGa WssYLxeaedrQqLidi5L8z4mr1KxyXhR7o4NCYxihpw1F0dofdYM/5jXiklIi944p8fvV 6dKjM9cvtpzTxi7ijEKcESkZOH7RhDPusTZvcD81qbecXn/eMsaMQJl+FOy/o3aIlNXJ orAzizYMTyjVJL7SpBuTDD5bxdmis09WIyc6wFJFAQKYXtTiAcjuRt2p1/9sB/pbczGk 6EasR8/yJ+bsJssJ0W3nacY6X+Q/PnfastXyTkGVU51zgYHfauZNFIC4J1tVXHKsatwW 7MJA==
X-Gm-Message-State: AOAM533IAOkTROYwut2wr9wnql3L3TYk2pXgymKiQDzkQuNkXfYLPR7Z wbVgWNcAfP7h8/sKntci81N8/FMSLgy+vBDOPiM=
X-Google-Smtp-Source: ABdhPJyNuUT/qh7JEPxOLTpKcs0JwBFUFMW9ODmYI1fXODq3HUc1Ny2nJqOOrMgWFY0OX9McbFfC4D879BGm21VNSdY=
X-Received: by 2002:a02:8e:: with SMTP id 136mr5988893jaa.144.1592607625590; Fri, 19 Jun 2020 16:00:25 -0700 (PDT)
MIME-Version: 1.0
References: <CAM4esxQBqfrz24riPQA_VGKcGp_TzW0pqb97KfFMtNdW9pUfDg@mail.gmail.com> <833A693C-62E6-4889-9954-FCE65A839A7C@eggert.org> <CAKcm_gPMO2DtqvKucqVw0zDjSniSOmFD4p1Tp4YLjr9WSWdEUw@mail.gmail.com> <CAJU8_nUN42gGmQof24XD9-EjXedyzcarDyRP8MGe1qW-BZ=+Aw@mail.gmail.com> <9cd91c24-c730-22a4-7aa0-baf61613b3ce@huitema.net> <f4922cdb59014202900de44cc5fea0ff@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQvwkTvpUcu6-+W5zWo22m-R1jvN7DcCpXfuw8Hb55qsw@mail.gmail.com> <95dd02c92b32472d9cab0dd47b98c637@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQxxXn27rZEY75-AsHD5VF0fqiV1VDyeSrzQ=-sM7JNCg@mail.gmail.com> <9c2e300c30f74d1794d11cf4334ea07b@usma1ex-dag1mb5.msg.corp.akamai.com> <2c40f3d9-fa40-9834-ac30-36bc9a1a6303@huitema.net> <CA+9kkMBQt001xOVgT=9G8YOOTOJ+9S=OWDwGEeYuKVm46Cq3iQ@mail.gmail.com> <3bb42dfc-17ba-c5c8-03f7-35428756b4c2@huitema.net>
In-Reply-To: <3bb42dfc-17ba-c5c8-03f7-35428756b4c2@huitema.net>
From: Martin Duke <martin.h.duke@gmail.com>
Date: Fri, 19 Jun 2020 16:00:14 -0700
Message-ID: <CAM4esxRWVRjVhxyYuuzwDGq_wfTjQHkY6KHG2rEPErO2aHXA0w@mail.gmail.com>
Subject: Re: Is the invariants draft really standards track?
To: Christian Huitema <huitema@huitema.net>
Cc: Ted Hardie <ted.ietf@gmail.com>,  "Lubashev, Igor" <ilubashe=40akamai.com@dmarc.ietf.org>, IETF QUIC WG <quic@ietf.org>,  Lars Eggert <lars@eggert.org>, Kyle Rose <krose@krose.org>,  Ian Swett <ianswett=40google.com@dmarc.ietf.org>, Jared Mauch <jared@puck.nether.net>
Content-Type: multipart/alternative; boundary="000000000000ca3e6705a877db1d"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/nYDbgz0nBBX0Kp5GGLbgGTZ8SZA>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 19 Jun 2020 23:00:31 -0000

--000000000000ca3e6705a877db1d
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

>  The problem of the DOS protection box is to drop as much traffic as
possible while still letting some good in, and also without causing
performance hits when there are no attacks. There are indeed some thorny
traffic management issues there. More information helps, and Ben's
suggestion to look at DOTS is certainly on point.

I am not sure how any of this can work when inbound 1RTT packets can come
with essentially random IP/port and CID and potentially be valid, given
there might have been a migration. Some of the QUIC-LB modes at least make
it hard to guess your way to a valid CID.

DOTS is an intriguing idea, but again I'm not sure what signatures one
could use to configure it.

On Fri, Jun 19, 2020 at 3:43 PM Christian Huitema <huitema@huitema.net>
wrote:

>
> On 6/19/2020 3:05 PM, Ted Hardie wrote:
>
> Hi Christian,
>
> A question in-line,
>
> On Fri, Jun 19, 2020 at 2:49 PM Christian Huitema <huitema@huitema.net>
> wrote:
>
>> When under DOS attack, you want to "minimize blowback", i.e., as much as
>> possible avoid generating packets in response to attack traffic. So, yes=
, a
>> server may choose to not send stateless resets to anyone when under atta=
ck;
>> in fact, my recommendation would be that a server SHOULD NOT send statel=
ess
>> resets to anyone when under attack.
>>
>> That said, Igor raised an interesting point about return traffic. It
>> would be very nice if DOS protection boxes could distinguish between
>> "validated traffic" that the server presumably intends to process, and
>> "unsolicited traffic" that will just consume resource. The box can then
>> reserve some share of the resource for validated traffic, and place the
>> rest of the traffic in a lower priority queue. Fine, but there needs to =
be
>> a test. The classic test is that incoming traffic is "validated" if the
>> protection box can match it with return traffic coming from the server -=
-
>> for some definition of matching.
>>
>> From that point of view, stateless reset is definitely not helpful. But
>> problematic traffic goes beyond that. The server will reply to a client'=
s
>> initial with a server's initial packet. Does that validate the response
>> traffic? OK, maybe the protection box can programmed to only validate
>> traffic if it sees 1RTT packets. But many servers will send 1-RTT packet=
s
>> as 0.5 RTT. Does that validate the response traffic?
>>
>> We might say that traffic is validated when the handshake is confirmed,
>> but the protection box does not understand the TLS handshake, it just se=
es
>> packet types and packet sizes. It cannot distinguish between 0.5RTT data
>> and 1RTT data, and thus the closest approximation of "validation" would =
be
>> seeing more than an initial window worth of traffic coming back from the
>> server. That does not sound great.
>>
>> On the other hand, things get much better if the server under attack can
>> adopt a defensive posture and help the DOS protection box do its job.
>> Suppose that a server can detect that it is under attack -- or be
>> explicitly configured so. The simplest defensive posture would be to (1)
>> disable stateless reset and (2) not send any 0.5RTT packet, including in
>> response to 0-RTT. The protection boxes can at that point take the 1-RTT
>> packets from the server as indicating validation.
>>
> Perhaps I'm misreading you here, but this sounds like you expect the
> protection boxes to be able to distinguish when servers see themselves as
> under DOS from when they don't, so that they can tell that the lack of
> 0.5RTT is an indication of an attack response.  Given distribution patter=
ns
> of DOS attacks, I'm struggling to see whether that will commonly be the
> case.
>
> Maybe I wasn't too clear. I don't think that DOS protection boxes are
> trying to take directions from servers. I also don't believe that DOS
> protection boxes have any practical means of distinguishing between 0.5 R=
TT
> and 1 RTT traffic. That might actually be a problem.
>
>
> You could, of course, always put handshake traffic into low-priority
> queues until you see the 1-RTT packets that validate the server's
> interest.  That would make 1-RTT traffic effectively a path signal of
> validation.
>
> Yes, that's more or less what I was considering. But of course it only
> works if the server refrains from sending 1RTT packets until the handshak=
e
> is confirmed. Otherwise, the protection boxes will have to count the numb=
er
> of 1RTT packets coming back from the server, or maybe the amount of data
> coming from the server, and only consider the traffic validated if number
> of packets and amount of data are larger than common values of IW. You
> might end up with 3 classes instead of 2 -- validated if 1RTT > IW,
> validation in progress if 1RTT see but < IW, not validated yet if no 1RTT
> seen.
>
> My concern with that is that using those low priority queues during the
> handshake phase seems likely to result in worse latency and increased ris=
k
> of packet loss (which can be tricky during that phase).  That seems a hea=
vy
> price to pay during non-attack times for better protection when under
> attack.
>
> The problem of the DOS protection box is to drop as much traffic as
> possible while still letting some good in, and also without causing
> performance hits when there are no attacks. There are indeed some thorny
> traffic management issues there. More information helps, and Ben's
> suggestion to look at DOTS is certainly on point.
>
> I am just doing a thought experiment so far. One immediate observation is
> that in the absence of other data, DOS protection designers are likely to
> look at packet types, and certainly reason about 1RTT packets versus long
> header packets. This is definitely an ossification risk.
>
> The other observation is that we have to be careful about the realism of
> thought experiments. If I remember correctly, the payload of the Blaster
> worm was something like:
>
>     On each bot
>
>         Open 256 threads
>
>             In each thread, loop on "GET some large page from the server"
>
> Reasoning about validated traffic would not stop that. But reasoning abou=
t
> validated traffic forces the attacker to disclose the "real" IP addresses
> of the attacking bots, which then enables a second line of defense.
>
> -- Christian Huitema
>
>
>
>
> Am I misunderstanding how this is distinguished?
>
> Clue appreciated,
>
> Ted
>
>
>
>> Maybe we should specify that.
>>
>> -- Christian Huitema
>> On 6/19/2020 1:42 PM, Lubashev, Igor wrote:
>>
>> > There is no need for servers to decrypt CIDs in QUIC-LB. Presumably th=
e
>> server has a lookup table for its CIDs.
>>
>>
>>
>> Sending a stateless reset in response to a junk packet would cost more
>> CPU than verifying CID integrity.  But, yes, a server may choose to not
>> send stateless resets to anyone when under attack.
>>
>>
>>
>>
>>
>> *From:* Martin Duke <martin.h.duke@gmail.com> <martin.h.duke@gmail.com>
>> *Sent:* Friday, June 19, 2020 2:44 PM
>>
>> >Unfortunately, Retry system protects only server's memory state and som=
e
>> CPU cycles spent on crypto.  (Servers still need to decrypt CID to decid=
e
>> it is invalid, and if the attacker is clever enough to establish one val=
id
>> connection and use that CID in a flood, the server will also be decrypti=
ng
>> packets.)
>>
>>
>>
>> There is no need for servers to decrypt CIDs in QUIC-LB. Presumably the
>> server has a lookup table for its CIDs.
>>
>>
>>
>> It is true that Retry Services (and indeed, the Retry concept as a whole=
)
>> does nothing to protect network capacity.
>>
>>
>>
>> On Fri, Jun 19, 2020 at 8:08 AM Lubashev, Igor <ilubashe@akamai.com>
>> wrote:
>>
>> It looks like
>> https://tools.ietf.org/html/draft-ietf-quic-load-balancers-02#section-5
>> <https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__tools.ietf.org_h=
tml_draft-2Dietf-2Dquic-2Dload-2Dbalancers-2D02-23section-2D5&d=3DDwMFaQ&c=
=3D96ZbZZcaMF4w0F4jpN6LZg&r=3DDjn3bQ5uNJDPM_2skfL3rW1tzcIxyjUZdn_m55KPmlo&m=
=3DeTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTKljvj-ys&s=3D1x7AN2a51X_zV5xSyK0uCL8v=
Z5cugD3n4lbFWZDqVQg&e=3D>
>> is an excellent discussion of Retry mechanics.  It definitely deserves a
>> reference from this manageability draft.
>>
>>
>>
>> The Retry mechanisms described in LB draft are all cooperating boxes, an=
d
>> servers must be aware of them.  Unfortunately, Retry system protects onl=
y
>> server's memory state and some CPU cycles spent on crypto.  (Servers sti=
ll
>> need to decrypt CID to decide it is invalid, and if the attacker is clev=
er
>> enough to establish one valid connection and use that CID in a flood, th=
e
>> server will also be decrypting packets.)  Retry does nothing to protect
>> network resources.
>>
>>
>>
>> The PR I opened (https://github.com/quicwg/ops-drafts/pull/94
>> <https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__github.com_quicw=
g_ops-2Ddrafts_pull_94&d=3DDwMFaQ&c=3D96ZbZZcaMF4w0F4jpN6LZg&r=3DDjn3bQ5uNJ=
DPM_2skfL3rW1tzcIxyjUZdn_m55KPmlo&m=3DeTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTKl=
jvj-ys&s=3DVpvhQ9n79rANIk3O5Sm7PZyToFQEennoPt9iqFpWbq8&e=3D>)
>> is about uncooperating devices that try to mitigate volumetric network
>> attacks.
>>
>>
>>
>>
>>
>> *From:* Martin Duke <martin.h.duke@gmail.com>
>> *Sent:* Wednesday, June 17, 2020 8:16 PM
>>
>> Hi Igor, you might want to check out the "Retry Services" bit of the
>> QUIC-LB draft. This has something to do with the DDoS use case you discu=
ss.
>>
>>
>>
>> On Wed, May 27, 2020 at 9:07 AM Lubashev, Igor <ilubashe@akamai.com>
>> wrote:
>>
>> I=E2=80=99m working on a manageability draft PR for this (how to rate li=
mit UDP
>> to reduce disruption to QUIC if you have to rate limit UDP).  ETA end of
>> the week (if I do not get pulled into something again).
>>
>>
>>
>> The relevant observation is that DDoS with UDP that is indistinguishable
>> from QUIC will happen.  UDP is already the most prevalent DDoS vector,
>> since it is easy for a compromised non-admin app to send a flood of huge
>> UDP packets (with TCP you get throttled by the congestion controller).  =
So
>> there WILL be DDoS protection devices out there to try to mitigate the
>> problem, possibly by observing both directions of the flow and deciding
>> whether a packet belongs to a valid flow or not.
>>
>>
>>
>> Since such middle boxes will be created, the more explicit and normative
>> Invariants are about what one can expect, the less such middle boxes may
>> decide for themselves.  For example (I did not think long about it), if
>> some elements of path validation could land into Invariants (roughly, =
=E2=80=9Cno
>> more than X packets/bytes can be sent on a new path w/o a return packet=
=E2=80=9D),
>> a DDoS middle box may use this fact and active connection migration migh=
t
>> still have a chance during an attack (NAT rebinding could be linked by D=
DoS
>> boxes to an old connection via unchanged CID).
>>
>>
>>
>>    - Igor
>>
>>
>>
>>
>>
>> *From:* Christian Huitema <huitema@huitema.net>
>> *Sent:* Wednesday, May 27, 2020 11:34 AM
>>
>> On 5/27/2020 8:28 AM, Kyle Rose wrote:
>>
>> On Wed, May 27, 2020 at 10:34 AM Ian Swett <ianswett=3D
>> 40google.com@dmarc.ietf.org> wrote:
>>
>> I was agreeing with MT, but I'm happy to see some more MUSTs added if
>> people feel that'd be helpful.
>>
>>
>>
>> Coincidentally, we were just talking about this internally at Akamai
>> yesterday. IMO, an invariants document isn't really helpful if it isn't
>> normative, and for it to be normative it (or a related practices doc for
>> operators) really needs to spell out clear boundaries for operators with
>> MUSTs..
>>
>>
>>
>> The example that came up yesterday was around operators filtering QUIC i=
n
>> the event of a DDoS: one recommendation based on some conversations goin=
g
>> back at least to Prague 2019 was to hash packets on 4-tuple and filter
>> those below a hash value chosen for a desired ingress limit instead of
>> doing what most operators do with UDP today, which is to cap UDP through=
put
>> and just drop packets randomly or tail drop.
>>
>> Interesting. Did they consider using the CID, or a fraction of it? This
>> looks entirely like the scenario for which we developed stateless retry.
>>
>>
>>
>> This recommendation certainly imposes some constraints on future protoco=
l
>> development that motivate new invariants: for instance, it would preclud=
e
>> sharding a connection across multiple source ports (not that there is
>> necessarily a reason to do this; it's just an example). But more
>> importantly, it goes beyond invariants: it's one among many practices
>> compatible with the current set of invariants, some reasonable and some
>> terrible.
>>
>> This would break the "preferred address" redirection. Preferred address
>> migration may or may not be spelled out in the invariants.
>>
>>
>>
>> Operators are going to do things to QUIC traffic, so it would be good to
>> offer them recommendations that are compatible with broad deployability.
>>
>>
>>
>> Yes, we do need the invariants for that.
>>
>> -- Christian Huitema
>>
>>

--000000000000ca3e6705a877db1d
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">&gt;=C2=A0

The problem of the DOS protection box is to drop as much traffic as possibl=
e while still letting some good in, and also without causing performance hi=
ts when there are no attacks. There are indeed some thorny traffic manageme=
nt issues there. More information helps, and Ben&#39;s suggestion to look a=
t DOTS is certainly on point.<div><br></div><div>I am not sure how any of t=
his can work when inbound=C2=A01RTT packets can come with essentially rando=
m IP/port and CID and potentially be valid, given there might have been a m=
igration. Some of the QUIC-LB modes at least make it hard to guess your way=
 to a valid=C2=A0CID.</div><div><br></div><div>DOTS is an intriguing idea, =
but again I&#39;m not sure what signatures one could use to configure it.</=
div></div><br><div class=3D"gmail_quote"><div dir=3D"ltr" class=3D"gmail_at=
tr">On Fri, Jun 19, 2020 at 3:43 PM Christian Huitema &lt;<a href=3D"mailto=
:huitema@huitema.net">huitema@huitema.net</a>&gt; wrote:<br></div><blockquo=
te class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px =
solid rgb(204,204,204);padding-left:1ex">
 =20
   =20
 =20
  <div>
    <p><br>
    </p>
    <div>On 6/19/2020 3:05 PM, Ted Hardie wrote:<br>
    </div>
    <blockquote type=3D"cite">
     =20
      <div dir=3D"ltr">
        <div dir=3D"ltr">
          <div>Hi Christian,</div>
          <div><br>
          </div>
          <div>A question in-line,<br>
          </div>
        </div>
        <br>
        <div class=3D"gmail_quote">
          <div dir=3D"ltr" class=3D"gmail_attr">On Fri, Jun 19, 2020 at 2:4=
9
            PM Christian Huitema &lt;<a href=3D"mailto:huitema@huitema.net"=
 target=3D"_blank">huitema@huitema.net</a>&gt;
            wrote:<br>
          </div>
          <blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8=
ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
            <div>
              <p>When under DOS attack, you want to &quot;minimize blowback=
&quot;,
                i.e., as much as possible avoid generating packets in
                response to attack traffic. So, yes, a server may choose
                to not send stateless resets to anyone when under
                attack; in fact, my recommendation would be that a
                server SHOULD NOT send stateless resets to anyone when
                under attack.</p>
              <p>That said, Igor raised an interesting point about
                return traffic. It would be very nice if DOS protection
                boxes could distinguish between &quot;validated traffic&quo=
t; that
                the server presumably intends to process, and
                &quot;unsolicited traffic&quot; that will just consume reso=
urce.
                The box can then reserve some share of the resource for
                validated traffic, and place the rest of the traffic in
                a lower priority queue. Fine, but there needs to be a
                test. The classic test is that incoming traffic is
                &quot;validated&quot; if the protection box can match it wi=
th
                return traffic coming from the server -- for some
                definition of matching.</p>
              <p>From that point of view, stateless reset is definitely
                not helpful. But problematic traffic goes beyond that.
                The server will reply to a client&#39;s initial with a
                server&#39;s initial packet. Does that validate the respons=
e
                traffic? OK, maybe the protection box can programmed to
                only validate traffic if it sees 1RTT packets. But many
                servers will send 1-RTT packets as 0.5 RTT. Does that
                validate the response traffic?</p>
              <p>We might say that traffic is validated when the
                handshake is confirmed, but the protection box does not
                understand the TLS handshake, it just sees packet types
                and packet sizes. It cannot distinguish between 0.5RTT
                data and 1RTT data, and thus the closest approximation
                of &quot;validation&quot; would be seeing more than an init=
ial
                window worth of traffic coming back from the server.
                That does not sound great.</p>
              <p>On the other hand, things get much better if the server
                under attack can adopt a defensive posture and help the
                DOS protection box do its job. Suppose that a server can
                detect that it is under attack -- or be explicitly
                configured so. The simplest defensive posture would be
                to (1) disable stateless reset and (2) not send any
                0.5RTT packet, including in response to 0-RTT. The
                protection boxes can at that point take the 1-RTT
                packets from the server as indicating validation.</p>
            </div>
          </blockquote>
          <div>Perhaps I&#39;m misreading you here, but this sounds like yo=
u
            expect the protection boxes to be able to distinguish when
            servers see themselves as under DOS from when they don&#39;t, s=
o
            that they can tell that the lack of 0.5RTT is an indication
            of an attack response.=C2=A0 Given distribution patterns of DOS
            attacks, I&#39;m struggling to see whether that will commonly b=
e
            the case.</div>
        </div>
      </div>
    </blockquote>
    <p>Maybe I wasn&#39;t too clear. I don&#39;t think that DOS protection =
boxes
      are trying to take directions from servers. I also don&#39;t believe
      that DOS protection boxes have any practical means of
      distinguishing between 0.5 RTT and 1 RTT traffic. That might
      actually be a problem.<br>
    </p>
    <blockquote type=3D"cite">
      <div dir=3D"ltr">
        <div class=3D"gmail_quote">
          <div><br>
          </div>
          <div>You could, of course, always put handshake traffic into
            low-priority queues until you see the 1-RTT packets that
            validate the server&#39;s interest.=C2=A0 That would make 1-RTT
            traffic effectively a path signal of validation.=C2=A0 </div>
        </div>
      </div>
    </blockquote>
    <p>Yes, that&#39;s more or less what I was considering. But of course i=
t
      only works if the server refrains from sending 1RTT packets until
      the handshake is confirmed. Otherwise, the protection boxes will
      have to count the number of 1RTT packets coming back from the
      server, or maybe the amount of data coming from the server, and
      only consider the traffic validated if number of packets and
      amount of data are larger than common values of IW. You might end
      up with 3 classes instead of 2 -- validated if 1RTT &gt; IW,
      validation in progress if 1RTT see but &lt; IW, not validated yet
      if no 1RTT seen.<br>
    </p>
    <blockquote type=3D"cite">
      <div dir=3D"ltr">
        <div class=3D"gmail_quote">
          <div>My concern with that is that using those low priority
            queues during the handshake phase seems likely to result in
            worse latency and increased risk of packet loss (which can
            be tricky during that phase).=C2=A0 That seems a heavy price to
            pay during non-attack times for better protection when under
            attack.</div>
        </div>
      </div>
    </blockquote>
    <p>The problem of the DOS protection box is to drop as much traffic
      as possible while still letting some good in, and also without
      causing performance hits when there are no attacks. There are
      indeed some thorny traffic management issues there. More
      information helps, and Ben&#39;s suggestion to look at DOTS is
      certainly on point.</p>
    <p>I am just doing a thought experiment so far. One immediate
      observation is that in the absence of other data, DOS protection
      designers are likely to look at packet types, and certainly reason
      about 1RTT packets versus long header packets. This is definitely
      an ossification risk.</p>
    <p>The other observation is that we have to be careful about the
      realism of thought experiments. If I remember correctly, the
      payload of the Blaster worm was something like:</p>
    <p>=C2=A0=C2=A0=C2=A0 On each bot</p>
    <p>=C2=A0=C2=A0=C2=A0 =C2=A0=C2=A0=C2=A0 Open 256 threads</p>
    <p>=C2=A0=C2=A0=C2=A0 =C2=A0=C2=A0=C2=A0 =C2=A0=C2=A0=C2=A0 In each thr=
ead, loop on &quot;GET some large page from the
      server&quot;</p>
    <p>Reasoning about validated traffic would not stop that. But
      reasoning about validated traffic forces the attacker to disclose
      the &quot;real&quot; IP addresses of the attacking bots, which then e=
nables
      a second line of defense.<br>
    </p>
    <p>-- Christian Huitema<br>
    </p>
    <p><br>
    </p>
    <p><br>
    </p>
    <blockquote type=3D"cite">
      <div dir=3D"ltr">
        <div class=3D"gmail_quote">
          <div><br>
          </div>
          <div>Am I misunderstanding how this is distinguished?</div>
          <div><br>
          </div>
          <div>Clue appreciated,</div>
          <div><br>
          </div>
          <div>Ted<br>
          </div>
          <div><br>
          </div>
          <div>=C2=A0</div>
          <blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8=
ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
            <div>
              <p>Maybe we should specify that.</p>
              <p>-- Christian Huitema<br>
              </p>
              <div>On 6/19/2020 1:42 PM, Lubashev, Igor wrote:<br>
              </div>
              <blockquote type=3D"cite">
                <div>
                  <p class=3D"MsoNormal">&gt; There is no need for servers
                    to decrypt CIDs in QUIC-LB. Presumably the server
                    has a lookup table for its CIDs.</p>
                  <p class=3D"MsoNormal">=C2=A0</p>
                  <p class=3D"MsoNormal">Sending a stateless reset in
                    response to a junk packet would cost more CPU than
                    verifying CID integrity.=C2=A0 But, yes, a server may
                    choose to not send stateless resets to anyone when
                    under attack.</p>
                  <p class=3D"MsoNormal">=C2=A0</p>
                  <p class=3D"MsoNormal">=C2=A0</p>
                  <div style=3D"border-color:currentcolor currentcolor curr=
entcolor blue;border-style:none none none solid;border-width:medium medium =
medium 1.5pt;padding:0in 0in 0in 4pt">
                    <div>
                      <div style=3D"border-color:rgb(225,225,225) currentco=
lor currentcolor;border-style:solid none none;border-width:1pt medium mediu=
m;padding:3pt 0in 0in">
                        <p class=3D"MsoNormal"><b>From:</b> Martin Duke <a =
href=3D"mailto:martin.h.duke@gmail.com" target=3D"_blank">&lt;martin.h.duke=
@gmail.com&gt;</a>
                          <br>
                          <b>Sent:</b> Friday, June 19, 2020 2:44 PM<br>
                          <br>
                        </p>
                      </div>
                    </div>
                    <div>
                      <p class=3D"MsoNormal">&gt;Unfortunately, Retry
                        system protects only server&#39;s memory state and
                        some CPU cycles spent on crypto.=C2=A0 (Servers sti=
ll
                        need to decrypt CID to decide it is invalid, and
                        if the attacker is clever enough to establish
                        one valid connection and use that CID in a
                        flood, the server will also be decrypting
                        packets.)=C2=A0=C2=A0</p>
                      <div>
                        <p class=3D"MsoNormal">=C2=A0</p>
                      </div>
                      <div>
                        <p class=3D"MsoNormal">There is no need for
                          servers to decrypt CIDs in QUIC-LB. Presumably
                          the server has a lookup table for its CIDs.</p>
                      </div>
                      <div>
                        <p class=3D"MsoNormal">=C2=A0</p>
                      </div>
                      <div>
                        <p class=3D"MsoNormal">It is true that Retry
                          Services (and indeed, the Retry concept as a
                          whole) does nothing to protect network
                          capacity.</p>
                      </div>
                    </div>
                    <p class=3D"MsoNormal">=C2=A0</p>
                    <div>
                      <div>
                        <p class=3D"MsoNormal">On Fri, Jun 19, 2020 at
                          8:08 AM Lubashev, Igor &lt;<a href=3D"mailto:ilub=
ashe@akamai.com" target=3D"_blank">ilubashe@akamai.com</a>&gt;
                          wrote:</p>
                      </div>
                      <blockquote style=3D"border-color:currentcolor curren=
tcolor currentcolor rgb(204,204,204);border-style:none none none solid;bord=
er-width:medium medium medium 1pt;padding:0in 0in 0in 6pt;margin-left:4.8pt=
;margin-right:0in">
                        <div>
                          <div>
                            <p class=3D"MsoNormal">It looks like <a href=3D=
"https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__tools.ietf.org_html=
_draft-2Dietf-2Dquic-2Dload-2Dbalancers-2D02-23section-2D5&amp;d=3DDwMFaQ&a=
mp;c=3D96ZbZZcaMF4w0F4jpN6LZg&amp;r=3DDjn3bQ5uNJDPM_2skfL3rW1tzcIxyjUZdn_m5=
5KPmlo&amp;m=3DeTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTKljvj-ys&amp;s=3D1x7AN2a5=
1X_zV5xSyK0uCL8vZ5cugD3n4lbFWZDqVQg&amp;e=3D" target=3D"_blank">
https://tools.ietf.org/html/draft-ietf-quic-load-balancers-02#section-5</a>
                              is an excellent discussion of Retry
                              mechanics.=C2=A0 It definitely deserves a
                              reference from this manageability draft.</p>
                            <p class=3D"MsoNormal">=C2=A0</p>
                            <p class=3D"MsoNormal">The Retry mechanisms
                              described in LB draft are all cooperating
                              boxes, and servers must be aware of them.=C2=
=A0
                              Unfortunately, Retry system protects only
                              server&#39;s memory state and some CPU cycles
                              spent on crypto.=C2=A0 (Servers still need to
                              decrypt CID to decide it is invalid, and
                              if the attacker is clever enough to
                              establish one valid connection and use
                              that CID in a flood, the server will also
                              be decrypting packets.)=C2=A0 Retry does
                              nothing to protect network resources.</p>
                            <p class=3D"MsoNormal">=C2=A0</p>
                            <p class=3D"MsoNormal">The PR I opened (<a href=
=3D"https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__github.com_quicw=
g_ops-2Ddrafts_pull_94&amp;d=3DDwMFaQ&amp;c=3D96ZbZZcaMF4w0F4jpN6LZg&amp;r=
=3DDjn3bQ5uNJDPM_2skfL3rW1tzcIxyjUZdn_m55KPmlo&amp;m=3DeTT-BoZ1fMitywKRVSxp=
U3js0lhO0qkspYTKljvj-ys&amp;s=3DVpvhQ9n79rANIk3O5Sm7PZyToFQEennoPt9iqFpWbq8=
&amp;e=3D" target=3D"_blank">https://github.com/quicwg/ops-drafts/pull/94</=
a>)
                              is about uncooperating devices that try to
                              mitigate volumetric network attacks.</p>
                            <p class=3D"MsoNormal">=C2=A0</p>
                            <p class=3D"MsoNormal">=C2=A0</p>
                            <div style=3D"border-color:currentcolor current=
color currentcolor blue;border-style:none none none solid;border-width:medi=
um medium medium 1.5pt;padding:0in 0in 0in 4pt">
                              <div>
                                <div style=3D"border-color:rgb(225,225,225)=
 currentcolor currentcolor;border-style:solid none none;border-width:1pt me=
dium medium;padding:3pt 0in 0in">
                                  <p class=3D"MsoNormal" style=3D"margin-bo=
ttom:12pt"><b>From:</b>
                                    Martin Duke &lt;<a href=3D"mailto:marti=
n.h.duke@gmail.com" target=3D"_blank">martin.h.duke@gmail.com</a>&gt;
                                    <br>
                                    <b>Sent:</b> Wednesday, June 17,
                                    2020 8:16 PM</p>
                                </div>
                              </div>
                              <div>
                                <p class=3D"MsoNormal">Hi Igor, you might
                                  want to check out the &quot;Retry Service=
s&quot;
                                  bit of the QUIC-LB draft. This has
                                  something to do with the DDoS use case
                                  you discuss.</p>
                              </div>
                              <p class=3D"MsoNormal">=C2=A0</p>
                              <div>
                                <div>
                                  <p class=3D"MsoNormal">On Wed, May 27,
                                    2020 at 9:07 AM Lubashev, Igor &lt;<a h=
ref=3D"mailto:ilubashe@akamai.com" target=3D"_blank">ilubashe@akamai.com</a=
>&gt;
                                    wrote:</p>
                                </div>
                                <blockquote style=3D"border-color:currentco=
lor currentcolor currentcolor rgb(204,204,204);border-style:none none none =
solid;border-width:medium medium medium 1pt;padding:0in 0in 0in 6pt;margin:=
5pt 0in 5pt 4.8pt">
                                  <div>
                                    <div>
                                      <p class=3D"MsoNormal">I=E2=80=99m wo=
rking
                                        on a manageability draft PR for
                                        this (how to rate limit UDP to
                                        reduce disruption to QUIC if you
                                        have to rate limit UDP).=C2=A0 ETA
                                        end of the week (if I do not get
                                        pulled into something again).</p>
                                      <p class=3D"MsoNormal">=C2=A0</p>
                                      <p class=3D"MsoNormal">The relevant
                                        observation is that DDoS with
                                        UDP that is indistinguishable
                                        from QUIC will happen.=C2=A0 UDP is
                                        already the most prevalent DDoS
                                        vector, since it is easy for a
                                        compromised non-admin app to
                                        send a flood of huge UDP packets
                                        (with TCP you get throttled by
                                        the congestion controller).=C2=A0 S=
o
                                        there WILL be DDoS protection
                                        devices out there to try to
                                        mitigate the problem, possibly
                                        by observing both directions of
                                        the flow and deciding whether a
                                        packet belongs to a valid flow
                                        or not.</p>
                                      <p class=3D"MsoNormal">=C2=A0</p>
                                      <p class=3D"MsoNormal">Since such
                                        middle boxes will be created,
                                        the more explicit and normative
                                        Invariants are about what one
                                        can expect, the less such middle
                                        boxes may decide for
                                        themselves.=C2=A0 For example (I di=
d
                                        not think long about it), if
                                        some elements of path validation
                                        could land into Invariants
                                        (roughly, =E2=80=9Cno more than X
                                        packets/bytes can be sent on a
                                        new path w/o a return packet=E2=80=
=9D),
                                        a DDoS middle box may use this
                                        fact and active connection
                                        migration might still have a
                                        chance during an attack (NAT
                                        rebinding could be linked by
                                        DDoS boxes to an old connection
                                        via unchanged CID).</p>
                                      <p class=3D"MsoNormal">=C2=A0</p>
                                      <ul type=3D"disc">
                                        <li class=3D"MsoNormal"> Igor</li>
                                      </ul>
                                      <p class=3D"MsoNormal">=C2=A0</p>
                                      <p class=3D"MsoNormal">=C2=A0</p>
                                      <div style=3D"border-color:currentcol=
or currentcolor currentcolor blue;border-style:none none none solid;border-=
width:medium medium medium 1.5pt;padding:0in 0in 0in 4pt">
                                        <p class=3D"MsoNormal" style=3D"mar=
gin-bottom:12pt"><b>From:</b>
                                          Christian Huitema &lt;<a href=3D"=
mailto:huitema@huitema.net" target=3D"_blank">huitema@huitema.net</a>&gt;
                                          <br>
                                          <b>Sent:</b> Wednesday, May
                                          27, 2020 11:34 AM</p>
                                        <div>
                                          <p class=3D"MsoNormal">On
                                            5/27/2020 8:28 AM, Kyle Rose
                                            wrote:</p>
                                        </div>
                                        <blockquote style=3D"margin-top:5pt=
;margin-bottom:5pt">
                                          <p class=3D"MsoNormal">On Wed,
                                            May 27, 2020 at 10:34 AM Ian
                                            Swett &lt;ianswett=3D<a href=3D=
"mailto:40google.com@dmarc.ietf.org" target=3D"_blank">40google.com@dmarc.i=
etf.org</a>&gt;
                                            wrote:</p>
                                          <div>
                                            <blockquote style=3D"border-col=
or:currentcolor currentcolor currentcolor rgb(204,204,204);border-style:non=
e none none solid;border-width:medium medium medium 1pt;padding:0in 0in 0in=
 6pt;margin:5pt 0in 5pt 4.8pt">
                                              <div>
                                                <p class=3D"MsoNormal">I
                                                  was agreeing with MT,
                                                  but I&#39;m happy to=C2=
=A0see
                                                  some more MUSTs added
                                                  if people feel that&#39;d
                                                  be helpful.</p>
                                              </div>
                                            </blockquote>
                                            <div>
                                              <p class=3D"MsoNormal"><span =
style=3D"font-size:12pt">=C2=A0</span></p>
                                              <div>
                                                <div>
                                                  <p class=3D"MsoNormal"><s=
pan style=3D"font-size:12pt">Coincidentally, we were just talking about thi=
s
                                                      internally at
                                                      Akamai yesterday.
                                                      IMO, an invariants
                                                      document isn&#39;t
                                                      really helpful if
                                                      it isn&#39;t
                                                      normative, and for
                                                      it to be normative
                                                      it (or a related
                                                      practices doc for
                                                      operators) really
                                                      needs to spell out
                                                      clear boundaries
                                                      for operators with
                                                      MUSTs..</span></p>
                                                </div>
                                                <div>
                                                  <p class=3D"MsoNormal"><s=
pan style=3D"font-size:12pt">=C2=A0</span></p>
                                                </div>
                                                <div>
                                                  <p class=3D"MsoNormal"><s=
pan style=3D"font-size:12pt">The example that came up yesterday was around
                                                      operators
                                                      filtering QUIC in
                                                      the event of a
                                                      DDoS: one
                                                      recommendation
                                                      based on some
                                                      conversations
                                                      going back at
                                                      least to Prague
                                                      2019 was to hash
                                                      packets on 4-tuple
                                                      and filter those
                                                      below a hash value
                                                      chosen for a
                                                      desired ingress
                                                      limit instead of
                                                      doing what most
                                                      operators do with
                                                      UDP today, which
                                                      is to cap UDP
                                                      throughput and
                                                      just drop packets
                                                      randomly or tail
                                                      drop.</span></p>
                                                </div>
                                              </div>
                                            </div>
                                          </div>
                                        </blockquote>
                                        <p>Interesting. Did they
                                          consider using the CID, or a
                                          fraction of it? This looks
                                          entirely like the scenario for
                                          which we developed stateless
                                          retry.</p>
                                        <blockquote style=3D"margin-top:5pt=
;margin-bottom:5pt">
                                          <div>
                                            <div>
                                              <div>
                                                <div>
                                                  <p class=3D"MsoNormal"><s=
pan style=3D"font-size:12pt">=C2=A0</span></p>
                                                </div>
                                                <div>
                                                  <p class=3D"MsoNormal"><s=
pan style=3D"font-size:12pt">This recommendation certainly imposes some
                                                      constraints on
                                                      future protocol
                                                      development that
                                                      motivate new
                                                      invariants: for
                                                      instance, it would
                                                      preclude sharding
                                                      a connection
                                                      across multiple
                                                      source ports (not
                                                      that there is
                                                      necessarily a
                                                      reason to do this;
                                                      it&#39;s just an
                                                      example). But more
                                                      importantly, it
                                                      goes beyond
                                                      invariants: it&#39;s
                                                      one among many
                                                      practices
                                                      compatible with
                                                      the current set of
                                                      invariants, some
                                                      reasonable and
                                                      some terrible.</span>=
</p>
                                                </div>
                                              </div>
                                            </div>
                                          </div>
                                        </blockquote>
                                        <p>This would break the
                                          &quot;preferred address&quot;
                                          redirection. Preferred address
                                          migration may or may not be
                                          spelled out in the invariants.</p=
>
                                        <blockquote style=3D"margin-top:5pt=
;margin-bottom:5pt">
                                          <div>
                                            <div>
                                              <div>
                                                <div>
                                                  <p class=3D"MsoNormal"><s=
pan style=3D"font-size:12pt">=C2=A0</span></p>
                                                </div>
                                                <div>
                                                  <p class=3D"MsoNormal"><s=
pan style=3D"font-size:12pt">Operators are going to do things to QUIC traff=
ic,
                                                      so it would be
                                                      good to offer them
                                                      recommendations
                                                      that are
                                                      compatible with
                                                      broad
                                                      deployability.</span>=
</p>
                                                </div>
                                              </div>
                                            </div>
                                          </div>
                                        </blockquote>
                                        <p>=C2=A0</p>
                                        <p>Yes, we do need the
                                          invariants for that.</p>
                                        <p>-- Christian Huitema</p>
                                      </div>
                                    </div>
                                  </div>
                                </blockquote>
                              </div>
                            </div>
                          </div>
                        </div>
                      </blockquote>
                    </div>
                  </div>
                </div>
              </blockquote>
            </div>
          </blockquote>
        </div>
      </div>
    </blockquote>
  </div>

</blockquote></div>

--000000000000ca3e6705a877db1d--


From nobody Fri Jun 19 16:51:49 2020
Return-Path: <huitema@huitema.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 7EB2A3A0F53 for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 16:51:47 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.798
X-Spam-Level: 
X-Spam-Status: No, score=-1.798 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, HTTPS_HTTP_MISMATCH=0.1, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=no autolearn_force=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 3ohY9gKA-2ly for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 16:51:44 -0700 (PDT)
Received: from mx36-out10.antispamcloud.com (mx36-out10.antispamcloud.com [209.126.121.30]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 7D71F3A0E17 for <quic@ietf.org>; Fri, 19 Jun 2020 16:51:44 -0700 (PDT)
Received: from xse20.mail2web.com ([66.113.196.20] helo=xse.mail2web.com) by mx114.antispamcloud.com with esmtp (Exim 4.92) (envelope-from <huitema@huitema.net>) id 1jmQnO-00016r-Pl for quic@ietf.org; Sat, 20 Jun 2020 01:51:44 +0200
Received: from xsmtp21.mail2web.com (unknown [10.100.68.60]) by xse.mail2web.com (Postfix) with ESMTPS id 49pbDr1pYJzD2nL for <quic@ietf.org>; Fri, 19 Jun 2020 16:51:40 -0700 (PDT)
Received: from [10.5.2.16] (helo=xmail06.myhosting.com) by xsmtp21.mail2web.com with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:256) (Exim 4.92) (envelope-from <huitema@huitema.net>) id 1jmQnM-0006rn-3i for quic@ietf.org; Fri, 19 Jun 2020 16:51:40 -0700
Received: (qmail 2724 invoked from network); 19 Jun 2020 23:51:39 -0000
Received: from unknown (HELO [192.168.1.107]) (Authenticated-user:_huitema@huitema.net@[172.58.43.64]) (envelope-sender <huitema@huitema.net>) by xmail06.myhosting.com (qmail-ldap-1.03) with ESMTPA for <jared@puck.nether.net>; 19 Jun 2020 23:51:39 -0000
To: Martin Duke <martin.h.duke@gmail.com>
Cc: Ted Hardie <ted.ietf@gmail.com>, "Lubashev, Igor" <ilubashe=40akamai.com@dmarc.ietf.org>, IETF QUIC WG <quic@ietf.org>, Lars Eggert <lars@eggert.org>, Kyle Rose <krose@krose.org>, Ian Swett <ianswett=40google.com@dmarc.ietf.org>, Jared Mauch <jared@puck.nether.net>
References: <CAM4esxQBqfrz24riPQA_VGKcGp_TzW0pqb97KfFMtNdW9pUfDg@mail.gmail.com> <833A693C-62E6-4889-9954-FCE65A839A7C@eggert.org> <CAKcm_gPMO2DtqvKucqVw0zDjSniSOmFD4p1Tp4YLjr9WSWdEUw@mail.gmail.com> <CAJU8_nUN42gGmQof24XD9-EjXedyzcarDyRP8MGe1qW-BZ=+Aw@mail.gmail.com> <9cd91c24-c730-22a4-7aa0-baf61613b3ce@huitema.net> <f4922cdb59014202900de44cc5fea0ff@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQvwkTvpUcu6-+W5zWo22m-R1jvN7DcCpXfuw8Hb55qsw@mail.gmail.com> <95dd02c92b32472d9cab0dd47b98c637@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQxxXn27rZEY75-AsHD5VF0fqiV1VDyeSrzQ=-sM7JNCg@mail.gmail.com> <9c2e300c30f74d1794d11cf4334ea07b@usma1ex-dag1mb5.msg.corp.akamai.com> <2c40f3d9-fa40-9834-ac30-36bc9a1a6303@huitema.net> <CA+9kkMBQt001xOVgT=9G8YOOTOJ+9S=OWDwGEeYuKVm46Cq3iQ@mail.gmail.com> <3bb42dfc-17ba-c5c8-03f7-35428756b4c2@huitema.net> <CAM4esxRWVRjVhxyYuuzwDGq_wfTjQHkY6KHG2rEPErO2aHXA0w@mail.gmail.com>
From: Christian Huitema <huitema@huitema.net>
Autocrypt: addr=huitema@huitema.net; prefer-encrypt=mutual; keydata= mDMEXtavGxYJKwYBBAHaRw8BAQdA1ou9A5MHTP9N3jfsWzlDZ+jPnQkusmc7sfLmWVz1Rmu0 J0NocmlzdGlhbiBIdWl0ZW1hIDxodWl0ZW1hQGh1aXRlbWEubmV0PoiWBBMWCAA+FiEEw3G4 Nwi4QEpAAXUUELAmqKBYtJQFAl7WrxsCGwMFCQlmAYAFCwkIBwIGFQoJCAsCBBYCAwECHgEC F4AACgkQELAmqKBYtJQbMwD/ebj/qnSbthC/5kD5DxZ/Ip0CGJw5QBz/+fJp3R8iAlsBAMjK r2tmyWyJz0CUkVG24WaR5EAJDvgwDv8h22U6QVkAuDgEXtavGxIKKwYBBAGXVQEFAQEHQJoM 6MUAIqpoqdCIiACiEynZf7nlJg2Eu0pXIhbUGONdAwEIB4h+BBgWCAAmFiEEw3G4Nwi4QEpA AXUUELAmqKBYtJQFAl7WrxsCGwwFCQlmAYAACgkQELAmqKBYtJRm2wD7BzeK5gEXSmBcBf0j BYdSaJcXNzx4yPLbP4GnUMAyl2cBAJzcsR4RkwO4dCRqM9CHpVJCwHtbUDJaa55//E0kp+gH
Subject: Re: Is the invariants draft really standards track?
Message-ID: <f9e2c611-bb4d-bc80-dfe3-e323a08bfc5b@huitema.net>
Date: Fri, 19 Jun 2020 16:51:39 -0700
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:68.0) Gecko/20100101 Thunderbird/68.9.0
MIME-Version: 1.0
In-Reply-To: <CAM4esxRWVRjVhxyYuuzwDGq_wfTjQHkY6KHG2rEPErO2aHXA0w@mail.gmail.com>
Content-Type: multipart/alternative; boundary="------------76EBE25B1A927C0BB99D6ABB"
Content-Language: en-US
X-Originating-IP: 66.113.196.20
X-Spampanel-Domain: xsmtpout.mail2web.com
X-Spampanel-Username: 66.113.196.20/32
Authentication-Results: antispamcloud.com; auth=pass smtp.auth=66.113.196.20/32@xsmtpout.mail2web.com
X-Spampanel-Outgoing-Class: unsure
X-Spampanel-Outgoing-Evidence: Combined (0.15)
X-Recommended-Action: accept
X-Filter-ID: Mvzo4OR0dZXEDF/gcnlw0f6LF1GdvkEexklpcFpSF5apSDasLI4SayDByyq9LIhVUZbR67CQ7/vm /hHDJU4RXkTNWdUk1Ol2OGx3IfrIJKywOmJyM1qr8uRnWBrbSAGDcnqpk5VeF3xR4kF6iVwRtbgN zB/4Jkrw1eDLcif59fuPvB4xtP0sauJTS+9eh3Bp+rYZvu7UEJiU3s27VgKHO7lwS3dBJTnTxDoD vBGGxph9w6EwXICYy0ePXtGEMhqrG0ZbIFW2KrxTvq8XOCo3J2U6UgOqKJ9sMwhVoOBGSAIboXtx P9OF0EfNs5TqNq2Yhy7LI0kfFnXdPP6btp4oBeJDeKRq5oPj2hFJhLx+qI3HlR3ootg7OlA3N5WN re/oppAGOX5cHTu1yz4pRT/9FGrxEaaKeSxe0Wrx6M4G5/WoLsdfEoJI0BNUQ4KpaNyNCwGqOUcw rXf55E8Tb8bmXq4yH8StrboPphDtmrtUkwkDMc9xayd+oZJo2heFY+g6kVWClPVvbW5lVyQanRxw 5rdY2rW50fd1ekaDpmIWc1Vmt3mnxMTQMQWbvBqEXskTQn6USYs98Imn+lZXe3dwYfgVB1xo6dCf BaU/iegBU8aHAjc56T3GRXf4QLCwDHpCLOlIZlwMa5gwaV7epgH0Aa5Q4MUtM8UyAp8ygf0yWkt3 abJVHw3Qw8OjNJfyX2Htqt+YCdgoyAPypW8k0Rbn9KzNjLvclnGzlTC8ZgkR3laMWpHinwUZhktg Vr2pl68EamsbOWjFhmHsIyPTDDMBbV2jZAOanSBpz6Rja2u/0jKQ4IgVsQW72rXuqOyd8ZppTZYh Llri23ia6HDqcuNUS0+FqboeKny9Q71ZpU7eHyLEZLCk58peHun779Lsjqfg7cTs80/2FnZg/IMs IAdedSzLrjsyfTPCYbMCLdmf5h2vfxw3Qvb2Glio5Cia/9Kfg4kJ0WtAYbrpe3OOAtQNb87OBHCz Hbokiue7PjVB1S6AQRz4SqXhOP5fdiQt7lu5Jm5nk4BSgYHOJJgUtm67rBRli6kULE5BQDZnPvvF VsQ=
X-Report-Abuse-To: spam@quarantine11.antispamcloud.com
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/-J_bzQZETwtPpWlB_AJ955prjCM>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 19 Jun 2020 23:51:48 -0000

This is a multi-part message in MIME format.
--------------76EBE25B1A927C0BB99D6ABB
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable


On 6/19/2020 4:00 PM, Martin Duke wrote:
> >=C2=A0 The problem of the DOS protection box is to drop as much traffi=
c as
> possible while still letting some good in, and also without causing
> performance hits when there are no attacks. There are indeed some
> thorny traffic management issues there. More information helps, and
> Ben's suggestion to look at DOTS is certainly on point.
>
> I am not sure how any of this can work when inbound=C2=A01RTT packets c=
an
> come with essentially random IP/port and CID and potentially be valid,
> given there might have been a migration. Some of the QUIC-LB modes at
> least make it hard to guess your way to a valid=C2=A0CID.

The DOS box does not have to worry about what kind of traffic is coming
in. It just has to open a context for the 5 tuple, and check whether it
sees 1-RTT packets coming back. And then maybe count the volume of 1-RTT
packets coming back.

The worry is that one of the bots might start a legitimate connection,
then disclose its five tuple to the rest of the botnet. The whole botnet
can then spoof the 5 tuple that was just pin-holed by the DOS box. A
simple "open-close" logic is thus not good enough. The DOS box must also
enforce some kind of rate limiting per 5 tuple.

Which also means that if a botnet can predict the 5 tuple used by a
legitimate connection and then spoof it, it can DOS it. Once you start
digging that particular rabbit hole, the joy never stops...

-- Christian Huitema


>
> DOTS is an intriguing idea, but again I'm not sure what signatures one
> could use to configure it.
>
> On Fri, Jun 19, 2020 at 3:43 PM Christian Huitema <huitema@huitema.net
> <mailto:huitema@huitema.net>> wrote:
>
>
>     On 6/19/2020 3:05 PM, Ted Hardie wrote:
>>     Hi Christian,
>>
>>     A question in-line,
>>
>>     On Fri, Jun 19, 2020 at 2:49 PM Christian Huitema
>>     <huitema@huitema.net <mailto:huitema@huitema.net>> wrote:
>>
>>         When under DOS attack, you want to "minimize blowback", i.e.,
>>         as much as possible avoid generating packets in response to
>>         attack traffic. So, yes, a server may choose to not send
>>         stateless resets to anyone when under attack; in fact, my
>>         recommendation would be that a server SHOULD NOT send
>>         stateless resets to anyone when under attack.
>>
>>         That said, Igor raised an interesting point about return
>>         traffic. It would be very nice if DOS protection boxes could
>>         distinguish between "validated traffic" that the server
>>         presumably intends to process, and "unsolicited traffic" that
>>         will just consume resource. The box can then reserve some
>>         share of the resource for validated traffic, and place the
>>         rest of the traffic in a lower priority queue. Fine, but
>>         there needs to be a test. The classic test is that incoming
>>         traffic is "validated" if the protection box can match it
>>         with return traffic coming from the server -- for some
>>         definition of matching.
>>
>>         From that point of view, stateless reset is definitely not
>>         helpful. But problematic traffic goes beyond that. The server
>>         will reply to a client's initial with a server's initial
>>         packet. Does that validate the response traffic? OK, maybe
>>         the protection box can programmed to only validate traffic if
>>         it sees 1RTT packets. But many servers will send 1-RTT
>>         packets as 0.5 RTT. Does that validate the response traffic?
>>
>>         We might say that traffic is validated when the handshake is
>>         confirmed, but the protection box does not understand the TLS
>>         handshake, it just sees packet types and packet sizes. It
>>         cannot distinguish between 0.5RTT data and 1RTT data, and
>>         thus the closest approximation of "validation" would be
>>         seeing more than an initial window worth of traffic coming
>>         back from the server. That does not sound great.
>>
>>         On the other hand, things get much better if the server under
>>         attack can adopt a defensive posture and help the DOS
>>         protection box do its job. Suppose that a server can detect
>>         that it is under attack -- or be explicitly configured so.
>>         The simplest defensive posture would be to (1) disable
>>         stateless reset and (2) not send any 0.5RTT packet, including
>>         in response to 0-RTT. The protection boxes can at that point
>>         take the 1-RTT packets from the server as indicating validatio=
n.
>>
>>     Perhaps I'm misreading you here, but this sounds like you expect
>>     the protection boxes to be able to distinguish when servers see
>>     themselves as under DOS from when they don't, so that they can
>>     tell that the lack of 0.5RTT is an indication of an attack
>>     response.=C2=A0 Given distribution patterns of DOS attacks, I'm
>>     struggling to see whether that will commonly be the case.
>
>     Maybe I wasn't too clear. I don't think that DOS protection boxes
>     are trying to take directions from servers. I also don't believe
>     that DOS protection boxes have any practical means of
>     distinguishing between 0.5 RTT and 1 RTT traffic. That might
>     actually be a problem.
>
>>
>>     You could, of course, always put handshake traffic into
>>     low-priority queues until you see the 1-RTT packets that validate
>>     the server's interest.=C2=A0 That would make 1-RTT traffic effecti=
vely
>>     a path signal of validation.=C2=A0
>
>     Yes, that's more or less what I was considering. But of course it
>     only works if the server refrains from sending 1RTT packets until
>     the handshake is confirmed. Otherwise, the protection boxes will
>     have to count the number of 1RTT packets coming back from the
>     server, or maybe the amount of data coming from the server, and
>     only consider the traffic validated if number of packets and
>     amount of data are larger than common values of IW. You might end
>     up with 3 classes instead of 2 -- validated if 1RTT > IW,
>     validation in progress if 1RTT see but < IW, not validated yet if
>     no 1RTT seen.
>
>>     My concern with that is that using those low priority queues
>>     during the handshake phase seems likely to result in worse
>>     latency and increased risk of packet loss (which can be tricky
>>     during that phase).=C2=A0 That seems a heavy price to pay during
>>     non-attack times for better protection when under attack.
>
>     The problem of the DOS protection box is to drop as much traffic
>     as possible while still letting some good in, and also without
>     causing performance hits when there are no attacks. There are
>     indeed some thorny traffic management issues there. More
>     information helps, and Ben's suggestion to look at DOTS is
>     certainly on point.
>
>     I am just doing a thought experiment so far. One immediate
>     observation is that in the absence of other data, DOS protection
>     designers are likely to look at packet types, and certainly reason
>     about 1RTT packets versus long header packets. This is definitely
>     an ossification risk.
>
>     The other observation is that we have to be careful about the
>     realism of thought experiments. If I remember correctly, the
>     payload of the Blaster worm was something like:
>
>     =C2=A0=C2=A0=C2=A0 On each bot
>
>     =C2=A0=C2=A0=C2=A0 =C2=A0=C2=A0=C2=A0 Open 256 threads
>
>     =C2=A0=C2=A0=C2=A0 =C2=A0=C2=A0=C2=A0 =C2=A0=C2=A0=C2=A0 In each th=
read, loop on "GET some large page from the
>     server"
>
>     Reasoning about validated traffic would not stop that. But
>     reasoning about validated traffic forces the attacker to disclose
>     the "real" IP addresses of the attacking bots, which then enables
>     a second line of defense.
>
>     -- Christian Huitema
>
>
>
>>
>>     Am I misunderstanding how this is distinguished?
>>
>>     Clue appreciated,
>>
>>     Ted
>>
>>     =C2=A0
>>
>>         Maybe we should specify that.
>>
>>         -- Christian Huitema
>>
>>         On 6/19/2020 1:42 PM, Lubashev, Igor wrote:
>>>
>>>         > There is no need for servers to decrypt CIDs in QUIC-LB.
>>>         Presumably the server has a lookup table for its CIDs.
>>>
>>>         =C2=A0
>>>
>>>         Sending a stateless reset in response to a junk packet would
>>>         cost more CPU than verifying CID integrity.=C2=A0 But, yes, a=

>>>         server may choose to not send stateless resets to anyone
>>>         when under attack.
>>>
>>>         =C2=A0
>>>
>>>         =C2=A0
>>>
>>>         *From:* Martin Duke <martin.h.duke@gmail.com>
>>>         <mailto:martin.h.duke@gmail.com>
>>>         *Sent:* Friday, June 19, 2020 2:44 PM
>>>
>>>         >Unfortunately, Retry system protects only server's memory
>>>         state and some CPU cycles spent on crypto.=C2=A0 (Servers sti=
ll
>>>         need to decrypt CID to decide it is invalid, and if the
>>>         attacker is clever enough to establish one valid connection
>>>         and use that CID in a flood, the server will also be
>>>         decrypting packets.)=C2=A0=C2=A0
>>>
>>>         =C2=A0
>>>
>>>         There is no need for servers to decrypt CIDs in QUIC-LB.
>>>         Presumably the server has a lookup table for its CIDs.
>>>
>>>         =C2=A0
>>>
>>>         It is true that Retry Services (and indeed, the Retry
>>>         concept as a whole) does nothing to protect network capacity.=

>>>
>>>         =C2=A0
>>>
>>>         On Fri, Jun 19, 2020 at 8:08 AM Lubashev, Igor
>>>         <ilubashe@akamai.com <mailto:ilubashe@akamai.com>> wrote:
>>>
>>>             It looks like
>>>             https://tools.ietf.org/html/draft-ietf-quic-load-balancer=
s-02#section-5
>>>             <https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__t=
ools.ietf.org_html_draft-2Dietf-2Dquic-2Dload-2Dbalancers-2D02-23section-=
2D5&d=3DDwMFaQ&c=3D96ZbZZcaMF4w0F4jpN6LZg&r=3DDjn3bQ5uNJDPM_2skfL3rW1tzcI=
xyjUZdn_m55KPmlo&m=3DeTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTKljvj-ys&s=3D1x7A=
N2a51X_zV5xSyK0uCL8vZ5cugD3n4lbFWZDqVQg&e=3D>
>>>             is an excellent discussion of Retry mechanics.=C2=A0 It
>>>             definitely deserves a reference from this manageability
>>>             draft.
>>>
>>>             =C2=A0
>>>
>>>             The Retry mechanisms described in LB draft are all
>>>             cooperating boxes, and servers must be aware of them.=C2=A0=

>>>             Unfortunately, Retry system protects only server's
>>>             memory state and some CPU cycles spent on crypto.=C2=A0
>>>             (Servers still need to decrypt CID to decide it is
>>>             invalid, and if the attacker is clever enough to
>>>             establish one valid connection and use that CID in a
>>>             flood, the server will also be decrypting packets.)=C2=A0=

>>>             Retry does nothing to protect network resources.
>>>
>>>             =C2=A0
>>>
>>>             The PR I opened
>>>             (https://github.com/quicwg/ops-drafts/pull/94
>>>             <https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__g=
ithub.com_quicwg_ops-2Ddrafts_pull_94&d=3DDwMFaQ&c=3D96ZbZZcaMF4w0F4jpN6L=
Zg&r=3DDjn3bQ5uNJDPM_2skfL3rW1tzcIxyjUZdn_m55KPmlo&m=3DeTT-BoZ1fMitywKRVS=
xpU3js0lhO0qkspYTKljvj-ys&s=3DVpvhQ9n79rANIk3O5Sm7PZyToFQEennoPt9iqFpWbq8=
&e=3D>)
>>>             is about uncooperating devices that try to mitigate
>>>             volumetric network attacks.
>>>
>>>             =C2=A0
>>>
>>>             =C2=A0
>>>
>>>             *From:* Martin Duke <martin.h.duke@gmail.com
>>>             <mailto:martin.h.duke@gmail.com>>
>>>             *Sent:* Wednesday, June 17, 2020 8:16 PM
>>>
>>>             Hi Igor, you might want to check out the "Retry
>>>             Services" bit of the QUIC-LB draft. This has something
>>>             to do with the DDoS use case you discuss.
>>>
>>>             =C2=A0
>>>
>>>             On Wed, May 27, 2020 at 9:07 AM Lubashev, Igor
>>>             <ilubashe@akamai.com <mailto:ilubashe@akamai.com>> wrote:=

>>>
>>>                 I=E2=80=99m working on a manageability draft PR for t=
his
>>>                 (how to rate limit UDP to reduce disruption to QUIC
>>>                 if you have to rate limit UDP).=C2=A0 ETA end of the =
week
>>>                 (if I do not get pulled into something again).
>>>
>>>                 =C2=A0
>>>
>>>                 The relevant observation is that DDoS with UDP that
>>>                 is indistinguishable from QUIC will happen.=C2=A0 UDP=
 is
>>>                 already the most prevalent DDoS vector, since it is
>>>                 easy for a compromised non-admin app to send a flood
>>>                 of huge UDP packets (with TCP you get throttled by
>>>                 the congestion controller).=C2=A0 So there WILL be DD=
oS
>>>                 protection devices out there to try to mitigate the
>>>                 problem, possibly by observing both directions of
>>>                 the flow and deciding whether a packet belongs to a
>>>                 valid flow or not.
>>>
>>>                 =C2=A0
>>>
>>>                 Since such middle boxes will be created, the more
>>>                 explicit and normative Invariants are about what one
>>>                 can expect, the less such middle boxes may decide
>>>                 for themselves.=C2=A0 For example (I did not think lo=
ng
>>>                 about it), if some elements of path validation could
>>>                 land into Invariants (roughly, =E2=80=9Cno more than =
X
>>>                 packets/bytes can be sent on a new path w/o a return
>>>                 packet=E2=80=9D), a DDoS middle box may use this fact=
 and
>>>                 active connection migration might still have a
>>>                 chance during an attack (NAT rebinding could be
>>>                 linked by DDoS boxes to an old connection via
>>>                 unchanged CID).
>>>
>>>                 =C2=A0
>>>
>>>                   * Igor
>>>
>>>                 =C2=A0
>>>
>>>                 =C2=A0
>>>
>>>                 *From:* Christian Huitema <huitema@huitema.net
>>>                 <mailto:huitema@huitema.net>>
>>>                 *Sent:* Wednesday, May 27, 2020 11:34 AM
>>>
>>>                 On 5/27/2020 8:28 AM, Kyle Rose wrote:
>>>
>>>                     On Wed, May 27, 2020 at 10:34 AM Ian Swett
>>>                     <ianswett=3D40google.com@dmarc.ietf.org
>>>                     <mailto:40google.com@dmarc.ietf.org>> wrote:
>>>
>>>                         I was agreeing with MT, but I'm happy to=C2=A0=
see
>>>                         some more MUSTs added if people feel that'd
>>>                         be helpful.
>>>
>>>                     =C2=A0
>>>
>>>                     Coincidentally, we were just talking about this
>>>                     internally at Akamai yesterday. IMO, an
>>>                     invariants document isn't really helpful if it
>>>                     isn't normative, and for it to be normative it
>>>                     (or a related practices doc for operators)
>>>                     really needs to spell out clear boundaries for
>>>                     operators with MUSTs..
>>>
>>>                     =C2=A0
>>>
>>>                     The example that came up yesterday was around
>>>                     operators filtering QUIC in the event of a DDoS:
>>>                     one recommendation based on some conversations
>>>                     going back at least to Prague 2019 was to hash
>>>                     packets on 4-tuple and filter those below a hash
>>>                     value chosen for a desired ingress limit instead
>>>                     of doing what most operators do with UDP today,
>>>                     which is to cap UDP throughput and just drop
>>>                     packets randomly or tail drop.
>>>
>>>                 Interesting. Did they consider using the CID, or a
>>>                 fraction of it? This looks entirely like the
>>>                 scenario for which we developed stateless retry.
>>>
>>>                     =C2=A0
>>>
>>>                     This recommendation certainly imposes some
>>>                     constraints on future protocol development that
>>>                     motivate new invariants: for instance, it would
>>>                     preclude sharding a connection across multiple
>>>                     source ports (not that there is necessarily a
>>>                     reason to do this; it's just an example). But
>>>                     more importantly, it goes beyond invariants:
>>>                     it's one among many practices compatible with
>>>                     the current set of invariants, some reasonable
>>>                     and some terrible.
>>>
>>>                 This would break the "preferred address"
>>>                 redirection. Preferred address migration may or may
>>>                 not be spelled out in the invariants.
>>>
>>>                     =C2=A0
>>>
>>>                     Operators are going to do things to QUIC
>>>                     traffic, so it would be good to offer them
>>>                     recommendations that are compatible with broad
>>>                     deployability.
>>>
>>>                 =C2=A0
>>>
>>>                 Yes, we do need the invariants for that.
>>>
>>>                 -- Christian Huitema
>>>

--------------76EBE25B1A927C0BB99D6ABB
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: 8bit

<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <p><br>
    </p>
    <div class="moz-cite-prefix">On 6/19/2020 4:00 PM, Martin Duke
      wrote:<br>
    </div>
    <blockquote type="cite"
cite="mid:CAM4esxRWVRjVhxyYuuzwDGq_wfTjQHkY6KHG2rEPErO2aHXA0w@mail.gmail.com">
      <meta http-equiv="content-type" content="text/html; charset=UTF-8">
      <div dir="ltr">&gt; 
        The problem of the DOS protection box is to drop as much traffic
        as possible while still letting some good in, and also without
        causing performance hits when there are no attacks. There are
        indeed some thorny traffic management issues there. More
        information helps, and Ben's suggestion to look at DOTS is
        certainly on point.
        <div><br>
        </div>
        <div>I am not sure how any of this can work when inbound 1RTT
          packets can come with essentially random IP/port and CID and
          potentially be valid, given there might have been a migration.
          Some of the QUIC-LB modes at least make it hard to guess your
          way to a valid CID.</div>
      </div>
    </blockquote>
    <p>The DOS box does not have to worry about what kind of traffic is
      coming in. It just has to open a context for the 5 tuple, and
      check whether it sees 1-RTT packets coming back. And then maybe
      count the volume of 1-RTT packets coming back.</p>
    <p>The worry is that one of the bots might start a legitimate
      connection, then disclose its five tuple to the rest of the
      botnet. The whole botnet can then spoof the 5 tuple that was just
      pin-holed by the DOS box. A simple "open-close" logic is thus not
      good enough. The DOS box must also enforce some kind of rate
      limiting per 5 tuple. <br>
    </p>
    <p>Which also means that if a botnet can predict the 5 tuple used by
      a legitimate connection and then spoof it, it can DOS it. Once you
      start digging that particular rabbit hole, the joy never stops...</p>
    <p>-- Christian Huitema<br>
    </p>
    <p><br>
    </p>
    <blockquote type="cite"
cite="mid:CAM4esxRWVRjVhxyYuuzwDGq_wfTjQHkY6KHG2rEPErO2aHXA0w@mail.gmail.com">
      <div dir="ltr">
        <div><br>
        </div>
        <div>DOTS is an intriguing idea, but again I'm not sure what
          signatures one could use to configure it.</div>
      </div>
      <br>
      <div class="gmail_quote">
        <div dir="ltr" class="gmail_attr">On Fri, Jun 19, 2020 at 3:43
          PM Christian Huitema &lt;<a href="mailto:huitema@huitema.net"
            moz-do-not-send="true">huitema@huitema.net</a>&gt; wrote:<br>
        </div>
        <blockquote class="gmail_quote" style="margin:0px 0px 0px
          0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
          <div>
            <p><br>
            </p>
            <div>On 6/19/2020 3:05 PM, Ted Hardie wrote:<br>
            </div>
            <blockquote type="cite">
              <div dir="ltr">
                <div dir="ltr">
                  <div>Hi Christian,</div>
                  <div><br>
                  </div>
                  <div>A question in-line,<br>
                  </div>
                </div>
                <br>
                <div class="gmail_quote">
                  <div dir="ltr" class="gmail_attr">On Fri, Jun 19, 2020
                    at 2:49 PM Christian Huitema &lt;<a
                      href="mailto:huitema@huitema.net" target="_blank"
                      moz-do-not-send="true">huitema@huitema.net</a>&gt;
                    wrote:<br>
                  </div>
                  <blockquote class="gmail_quote" style="margin:0px 0px
                    0px 0.8ex;border-left:1px solid
                    rgb(204,204,204);padding-left:1ex">
                    <div>
                      <p>When under DOS attack, you want to "minimize
                        blowback", i.e., as much as possible avoid
                        generating packets in response to attack
                        traffic. So, yes, a server may choose to not
                        send stateless resets to anyone when under
                        attack; in fact, my recommendation would be that
                        a server SHOULD NOT send stateless resets to
                        anyone when under attack.</p>
                      <p>That said, Igor raised an interesting point
                        about return traffic. It would be very nice if
                        DOS protection boxes could distinguish between
                        "validated traffic" that the server presumably
                        intends to process, and "unsolicited traffic"
                        that will just consume resource. The box can
                        then reserve some share of the resource for
                        validated traffic, and place the rest of the
                        traffic in a lower priority queue. Fine, but
                        there needs to be a test. The classic test is
                        that incoming traffic is "validated" if the
                        protection box can match it with return traffic
                        coming from the server -- for some definition of
                        matching.</p>
                      <p>From that point of view, stateless reset is
                        definitely not helpful. But problematic traffic
                        goes beyond that. The server will reply to a
                        client's initial with a server's initial packet.
                        Does that validate the response traffic? OK,
                        maybe the protection box can programmed to only
                        validate traffic if it sees 1RTT packets. But
                        many servers will send 1-RTT packets as 0.5 RTT.
                        Does that validate the response traffic?</p>
                      <p>We might say that traffic is validated when the
                        handshake is confirmed, but the protection box
                        does not understand the TLS handshake, it just
                        sees packet types and packet sizes. It cannot
                        distinguish between 0.5RTT data and 1RTT data,
                        and thus the closest approximation of
                        "validation" would be seeing more than an
                        initial window worth of traffic coming back from
                        the server. That does not sound great.</p>
                      <p>On the other hand, things get much better if
                        the server under attack can adopt a defensive
                        posture and help the DOS protection box do its
                        job. Suppose that a server can detect that it is
                        under attack -- or be explicitly configured so.
                        The simplest defensive posture would be to (1)
                        disable stateless reset and (2) not send any
                        0.5RTT packet, including in response to 0-RTT.
                        The protection boxes can at that point take the
                        1-RTT packets from the server as indicating
                        validation.</p>
                    </div>
                  </blockquote>
                  <div>Perhaps I'm misreading you here, but this sounds
                    like you expect the protection boxes to be able to
                    distinguish when servers see themselves as under DOS
                    from when they don't, so that they can tell that the
                    lack of 0.5RTT is an indication of an attack
                    response.  Given distribution patterns of DOS
                    attacks, I'm struggling to see whether that will
                    commonly be the case.</div>
                </div>
              </div>
            </blockquote>
            <p>Maybe I wasn't too clear. I don't think that DOS
              protection boxes are trying to take directions from
              servers. I also don't believe that DOS protection boxes
              have any practical means of distinguishing between 0.5 RTT
              and 1 RTT traffic. That might actually be a problem.<br>
            </p>
            <blockquote type="cite">
              <div dir="ltr">
                <div class="gmail_quote">
                  <div><br>
                  </div>
                  <div>You could, of course, always put handshake
                    traffic into low-priority queues until you see the
                    1-RTT packets that validate the server's interest. 
                    That would make 1-RTT traffic effectively a path
                    signal of validation.  </div>
                </div>
              </div>
            </blockquote>
            <p>Yes, that's more or less what I was considering. But of
              course it only works if the server refrains from sending
              1RTT packets until the handshake is confirmed. Otherwise,
              the protection boxes will have to count the number of 1RTT
              packets coming back from the server, or maybe the amount
              of data coming from the server, and only consider the
              traffic validated if number of packets and amount of data
              are larger than common values of IW. You might end up with
              3 classes instead of 2 -- validated if 1RTT &gt; IW,
              validation in progress if 1RTT see but &lt; IW, not
              validated yet if no 1RTT seen.<br>
            </p>
            <blockquote type="cite">
              <div dir="ltr">
                <div class="gmail_quote">
                  <div>My concern with that is that using those low
                    priority queues during the handshake phase seems
                    likely to result in worse latency and increased risk
                    of packet loss (which can be tricky during that
                    phase).  That seems a heavy price to pay during
                    non-attack times for better protection when under
                    attack.</div>
                </div>
              </div>
            </blockquote>
            <p>The problem of the DOS protection box is to drop as much
              traffic as possible while still letting some good in, and
              also without causing performance hits when there are no
              attacks. There are indeed some thorny traffic management
              issues there. More information helps, and Ben's suggestion
              to look at DOTS is certainly on point.</p>
            <p>I am just doing a thought experiment so far. One
              immediate observation is that in the absence of other
              data, DOS protection designers are likely to look at
              packet types, and certainly reason about 1RTT packets
              versus long header packets. This is definitely an
              ossification risk.</p>
            <p>The other observation is that we have to be careful about
              the realism of thought experiments. If I remember
              correctly, the payload of the Blaster worm was something
              like:</p>
            <p>    On each bot</p>
            <p>        Open 256 threads</p>
            <p>            In each thread, loop on "GET some large page
              from the server"</p>
            <p>Reasoning about validated traffic would not stop that.
              But reasoning about validated traffic forces the attacker
              to disclose the "real" IP addresses of the attacking bots,
              which then enables a second line of defense.<br>
            </p>
            <p>-- Christian Huitema<br>
            </p>
            <p><br>
            </p>
            <p><br>
            </p>
            <blockquote type="cite">
              <div dir="ltr">
                <div class="gmail_quote">
                  <div><br>
                  </div>
                  <div>Am I misunderstanding how this is distinguished?</div>
                  <div><br>
                  </div>
                  <div>Clue appreciated,</div>
                  <div><br>
                  </div>
                  <div>Ted<br>
                  </div>
                  <div><br>
                  </div>
                  <div> </div>
                  <blockquote class="gmail_quote" style="margin:0px 0px
                    0px 0.8ex;border-left:1px solid
                    rgb(204,204,204);padding-left:1ex">
                    <div>
                      <p>Maybe we should specify that.</p>
                      <p>-- Christian Huitema<br>
                      </p>
                      <div>On 6/19/2020 1:42 PM, Lubashev, Igor wrote:<br>
                      </div>
                      <blockquote type="cite">
                        <div>
                          <p class="MsoNormal">&gt; There is no need for
                            servers to decrypt CIDs in QUIC-LB.
                            Presumably the server has a lookup table for
                            its CIDs.</p>
                          <p class="MsoNormal"> </p>
                          <p class="MsoNormal">Sending a stateless reset
                            in response to a junk packet would cost more
                            CPU than verifying CID integrity.  But, yes,
                            a server may choose to not send stateless
                            resets to anyone when under attack.</p>
                          <p class="MsoNormal"> </p>
                          <p class="MsoNormal"> </p>
                          <div style="border-color:currentcolor
                            currentcolor currentcolor
                            blue;border-style:none none none
                            solid;border-width:medium medium medium
                            1.5pt;padding:0in 0in 0in 4pt">
                            <div>
                              <div style="border-color:rgb(225,225,225)
                                currentcolor
                                currentcolor;border-style:solid none
                                none;border-width:1pt medium
                                medium;padding:3pt 0in 0in">
                                <p class="MsoNormal"><b>From:</b> Martin
                                  Duke <a
                                    href="mailto:martin.h.duke@gmail.com"
                                    target="_blank"
                                    moz-do-not-send="true">&lt;martin.h.duke@gmail.com&gt;</a>
                                  <br>
                                  <b>Sent:</b> Friday, June 19, 2020
                                  2:44 PM<br>
                                  <br>
                                </p>
                              </div>
                            </div>
                            <div>
                              <p class="MsoNormal">&gt;Unfortunately,
                                Retry system protects only server's
                                memory state and some CPU cycles spent
                                on crypto.  (Servers still need to
                                decrypt CID to decide it is invalid, and
                                if the attacker is clever enough to
                                establish one valid connection and use
                                that CID in a flood, the server will
                                also be decrypting packets.)  </p>
                              <div>
                                <p class="MsoNormal"> </p>
                              </div>
                              <div>
                                <p class="MsoNormal">There is no need
                                  for servers to decrypt CIDs in
                                  QUIC-LB. Presumably the server has a
                                  lookup table for its CIDs.</p>
                              </div>
                              <div>
                                <p class="MsoNormal"> </p>
                              </div>
                              <div>
                                <p class="MsoNormal">It is true that
                                  Retry Services (and indeed, the Retry
                                  concept as a whole) does nothing to
                                  protect network capacity.</p>
                              </div>
                            </div>
                            <p class="MsoNormal"> </p>
                            <div>
                              <div>
                                <p class="MsoNormal">On Fri, Jun 19,
                                  2020 at 8:08 AM Lubashev, Igor &lt;<a
                                    href="mailto:ilubashe@akamai.com"
                                    target="_blank"
                                    moz-do-not-send="true">ilubashe@akamai.com</a>&gt;
                                  wrote:</p>
                              </div>
                              <blockquote
                                style="border-color:currentcolor
                                currentcolor currentcolor
                                rgb(204,204,204);border-style:none none
                                none solid;border-width:medium medium
                                medium 1pt;padding:0in 0in 0in
                                6pt;margin-left:4.8pt;margin-right:0in">
                                <div>
                                  <div>
                                    <p class="MsoNormal">It looks like <a
href="https://urldefense.proofpoint.com/v2/url?u=https-3A__tools.ietf.org_html_draft-2Dietf-2Dquic-2Dload-2Dbalancers-2D02-23section-2D5&amp;d=DwMFaQ&amp;c=96ZbZZcaMF4w0F4jpN6LZg&amp;r=Djn3bQ5uNJDPM_2skfL3rW1tzcIxyjUZdn_m55KPmlo&amp;m=eTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTKljvj-ys&amp;s=1x7AN2a51X_zV5xSyK0uCL8vZ5cugD3n4lbFWZDqVQg&amp;e="
                                        target="_blank"
                                        moz-do-not-send="true">
https://tools.ietf.org/html/draft-ietf-quic-load-balancers-02#section-5</a>
                                      is an excellent discussion of
                                      Retry mechanics.  It definitely
                                      deserves a reference from this
                                      manageability draft.</p>
                                    <p class="MsoNormal"> </p>
                                    <p class="MsoNormal">The Retry
                                      mechanisms described in LB draft
                                      are all cooperating boxes, and
                                      servers must be aware of them. 
                                      Unfortunately, Retry system
                                      protects only server's memory
                                      state and some CPU cycles spent on
                                      crypto.  (Servers still need to
                                      decrypt CID to decide it is
                                      invalid, and if the attacker is
                                      clever enough to establish one
                                      valid connection and use that CID
                                      in a flood, the server will also
                                      be decrypting packets.)  Retry
                                      does nothing to protect network
                                      resources.</p>
                                    <p class="MsoNormal"> </p>
                                    <p class="MsoNormal">The PR I opened
                                      (<a
href="https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_quicwg_ops-2Ddrafts_pull_94&amp;d=DwMFaQ&amp;c=96ZbZZcaMF4w0F4jpN6LZg&amp;r=Djn3bQ5uNJDPM_2skfL3rW1tzcIxyjUZdn_m55KPmlo&amp;m=eTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTKljvj-ys&amp;s=VpvhQ9n79rANIk3O5Sm7PZyToFQEennoPt9iqFpWbq8&amp;e="
                                        target="_blank"
                                        moz-do-not-send="true">https://github.com/quicwg/ops-drafts/pull/94</a>)
                                      is about uncooperating devices
                                      that try to mitigate volumetric
                                      network attacks.</p>
                                    <p class="MsoNormal"> </p>
                                    <p class="MsoNormal"> </p>
                                    <div
                                      style="border-color:currentcolor
                                      currentcolor currentcolor
                                      blue;border-style:none none none
                                      solid;border-width:medium medium
                                      medium 1.5pt;padding:0in 0in 0in
                                      4pt">
                                      <div>
                                        <div
                                          style="border-color:rgb(225,225,225)
                                          currentcolor
                                          currentcolor;border-style:solid
                                          none none;border-width:1pt
                                          medium medium;padding:3pt 0in
                                          0in">
                                          <p class="MsoNormal"
                                            style="margin-bottom:12pt"><b>From:</b>
                                            Martin Duke &lt;<a
                                              href="mailto:martin.h.duke@gmail.com"
                                              target="_blank"
                                              moz-do-not-send="true">martin.h.duke@gmail.com</a>&gt;
                                            <br>
                                            <b>Sent:</b> Wednesday, June
                                            17, 2020 8:16 PM</p>
                                        </div>
                                      </div>
                                      <div>
                                        <p class="MsoNormal">Hi Igor,
                                          you might want to check out
                                          the "Retry Services" bit of
                                          the QUIC-LB draft. This has
                                          something to do with the DDoS
                                          use case you discuss.</p>
                                      </div>
                                      <p class="MsoNormal"> </p>
                                      <div>
                                        <div>
                                          <p class="MsoNormal">On Wed,
                                            May 27, 2020 at 9:07 AM
                                            Lubashev, Igor &lt;<a
                                              href="mailto:ilubashe@akamai.com"
                                              target="_blank"
                                              moz-do-not-send="true">ilubashe@akamai.com</a>&gt;
                                            wrote:</p>
                                        </div>
                                        <blockquote
                                          style="border-color:currentcolor
                                          currentcolor currentcolor
                                          rgb(204,204,204);border-style:none
                                          none none
                                          solid;border-width:medium
                                          medium medium 1pt;padding:0in
                                          0in 0in 6pt;margin:5pt 0in 5pt
                                          4.8pt">
                                          <div>
                                            <div>
                                              <p class="MsoNormal">I’m
                                                working on a
                                                manageability draft PR
                                                for this (how to rate
                                                limit UDP to reduce
                                                disruption to QUIC if
                                                you have to rate limit
                                                UDP).  ETA end of the
                                                week (if I do not get
                                                pulled into something
                                                again).</p>
                                              <p class="MsoNormal"> </p>
                                              <p class="MsoNormal">The
                                                relevant observation is
                                                that DDoS with UDP that
                                                is indistinguishable
                                                from QUIC will happen. 
                                                UDP is already the most
                                                prevalent DDoS vector,
                                                since it is easy for a
                                                compromised non-admin
                                                app to send a flood of
                                                huge UDP packets (with
                                                TCP you get throttled by
                                                the congestion
                                                controller).  So there
                                                WILL be DDoS protection
                                                devices out there to try
                                                to mitigate the problem,
                                                possibly by observing
                                                both directions of the
                                                flow and deciding
                                                whether a packet belongs
                                                to a valid flow or not.</p>
                                              <p class="MsoNormal"> </p>
                                              <p class="MsoNormal">Since
                                                such middle boxes will
                                                be created, the more
                                                explicit and normative
                                                Invariants are about
                                                what one can expect, the
                                                less such middle boxes
                                                may decide for
                                                themselves.  For example
                                                (I did not think long
                                                about it), if some
                                                elements of path
                                                validation could land
                                                into Invariants
                                                (roughly, “no more than
                                                X packets/bytes can be
                                                sent on a new path w/o a
                                                return packet”), a DDoS
                                                middle box may use this
                                                fact and active
                                                connection migration
                                                might still have a
                                                chance during an attack
                                                (NAT rebinding could be
                                                linked by DDoS boxes to
                                                an old connection via
                                                unchanged CID).</p>
                                              <p class="MsoNormal"> </p>
                                              <ul type="disc">
                                                <li class="MsoNormal">
                                                  Igor</li>
                                              </ul>
                                              <p class="MsoNormal"> </p>
                                              <p class="MsoNormal"> </p>
                                              <div
                                                style="border-color:currentcolor
                                                currentcolor
                                                currentcolor
                                                blue;border-style:none
                                                none none
                                                solid;border-width:medium
                                                medium medium
                                                1.5pt;padding:0in 0in
                                                0in 4pt">
                                                <p class="MsoNormal"
                                                  style="margin-bottom:12pt"><b>From:</b>
                                                  Christian Huitema &lt;<a
href="mailto:huitema@huitema.net" target="_blank" moz-do-not-send="true">huitema@huitema.net</a>&gt;
                                                  <br>
                                                  <b>Sent:</b>
                                                  Wednesday, May 27,
                                                  2020 11:34 AM</p>
                                                <div>
                                                  <p class="MsoNormal">On
                                                    5/27/2020 8:28 AM,
                                                    Kyle Rose wrote:</p>
                                                </div>
                                                <blockquote
                                                  style="margin-top:5pt;margin-bottom:5pt">
                                                  <p class="MsoNormal">On
                                                    Wed, May 27, 2020 at
                                                    10:34 AM Ian Swett
                                                    &lt;ianswett=<a
                                                      href="mailto:40google.com@dmarc.ietf.org"
                                                      target="_blank"
                                                      moz-do-not-send="true">40google.com@dmarc.ietf.org</a>&gt;
                                                    wrote:</p>
                                                  <div>
                                                    <blockquote
                                                      style="border-color:currentcolor
                                                      currentcolor
                                                      currentcolor
                                                      rgb(204,204,204);border-style:none
                                                      none none
                                                      solid;border-width:medium
                                                      medium medium
                                                      1pt;padding:0in
                                                      0in 0in
                                                      6pt;margin:5pt 0in
                                                      5pt 4.8pt">
                                                      <div>
                                                        <p
                                                          class="MsoNormal">I
                                                          was agreeing
                                                          with MT, but
                                                          I'm happy
                                                          to see some
                                                          more MUSTs
                                                          added if
                                                          people feel
                                                          that'd be
                                                          helpful.</p>
                                                      </div>
                                                    </blockquote>
                                                    <div>
                                                      <p
                                                        class="MsoNormal"><span
style="font-size:12pt"> </span></p>
                                                      <div>
                                                        <div>
                                                          <p
                                                          class="MsoNormal"><span
style="font-size:12pt">Coincidentally, we were just talking about this
                                                          internally at
                                                          Akamai
                                                          yesterday.
                                                          IMO, an
                                                          invariants
                                                          document isn't
                                                          really helpful
                                                          if it isn't
                                                          normative, and
                                                          for it to be
                                                          normative it
                                                          (or a related
                                                          practices doc
                                                          for operators)
                                                          really needs
                                                          to spell out
                                                          clear
                                                          boundaries for
                                                          operators with
                                                          MUSTs..</span></p>
                                                        </div>
                                                        <div>
                                                          <p
                                                          class="MsoNormal"><span
style="font-size:12pt"> </span></p>
                                                        </div>
                                                        <div>
                                                          <p
                                                          class="MsoNormal"><span
style="font-size:12pt">The example that came up yesterday was around
                                                          operators
                                                          filtering QUIC
                                                          in the event
                                                          of a DDoS: one
                                                          recommendation
                                                          based on some
                                                          conversations
                                                          going back at
                                                          least to
                                                          Prague 2019
                                                          was to hash
                                                          packets on
                                                          4-tuple and
                                                          filter those
                                                          below a hash
                                                          value chosen
                                                          for a desired
                                                          ingress limit
                                                          instead of
                                                          doing what
                                                          most operators
                                                          do with UDP
                                                          today, which
                                                          is to cap UDP
                                                          throughput and
                                                          just drop
                                                          packets
                                                          randomly or
                                                          tail drop.</span></p>
                                                        </div>
                                                      </div>
                                                    </div>
                                                  </div>
                                                </blockquote>
                                                <p>Interesting. Did they
                                                  consider using the
                                                  CID, or a fraction of
                                                  it? This looks
                                                  entirely like the
                                                  scenario for which we
                                                  developed stateless
                                                  retry.</p>
                                                <blockquote
                                                  style="margin-top:5pt;margin-bottom:5pt">
                                                  <div>
                                                    <div>
                                                      <div>
                                                        <div>
                                                          <p
                                                          class="MsoNormal"><span
style="font-size:12pt"> </span></p>
                                                        </div>
                                                        <div>
                                                          <p
                                                          class="MsoNormal"><span
style="font-size:12pt">This recommendation certainly imposes some
                                                          constraints on
                                                          future
                                                          protocol
                                                          development
                                                          that motivate
                                                          new
                                                          invariants:
                                                          for instance,
                                                          it would
                                                          preclude
                                                          sharding a
                                                          connection
                                                          across
                                                          multiple
                                                          source ports
                                                          (not that
                                                          there is
                                                          necessarily a
                                                          reason to do
                                                          this; it's
                                                          just an
                                                          example). But
                                                          more
                                                          importantly,
                                                          it goes beyond
                                                          invariants:
                                                          it's one among
                                                          many practices
                                                          compatible
                                                          with the
                                                          current set of
                                                          invariants,
                                                          some
                                                          reasonable and
                                                          some terrible.</span></p>
                                                        </div>
                                                      </div>
                                                    </div>
                                                  </div>
                                                </blockquote>
                                                <p>This would break the
                                                  "preferred address"
                                                  redirection. Preferred
                                                  address migration may
                                                  or may not be spelled
                                                  out in the invariants.</p>
                                                <blockquote
                                                  style="margin-top:5pt;margin-bottom:5pt">
                                                  <div>
                                                    <div>
                                                      <div>
                                                        <div>
                                                          <p
                                                          class="MsoNormal"><span
style="font-size:12pt"> </span></p>
                                                        </div>
                                                        <div>
                                                          <p
                                                          class="MsoNormal"><span
style="font-size:12pt">Operators are going to do things to QUIC traffic,
                                                          so it would be
                                                          good to offer
                                                          them
                                                          recommendations
                                                          that are
                                                          compatible
                                                          with broad
                                                          deployability.</span></p>
                                                        </div>
                                                      </div>
                                                    </div>
                                                  </div>
                                                </blockquote>
                                                <p> </p>
                                                <p>Yes, we do need the
                                                  invariants for that.</p>
                                                <p>-- Christian Huitema</p>
                                              </div>
                                            </div>
                                          </div>
                                        </blockquote>
                                      </div>
                                    </div>
                                  </div>
                                </div>
                              </blockquote>
                            </div>
                          </div>
                        </div>
                      </blockquote>
                    </div>
                  </blockquote>
                </div>
              </div>
            </blockquote>
          </div>
        </blockquote>
      </div>
    </blockquote>
  </body>
</html>

--------------76EBE25B1A927C0BB99D6ABB--


From nobody Fri Jun 19 17:56:21 2020
Return-Path: <martin.h.duke@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 685F23A0FA2 for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 17:56:19 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.997
X-Spam-Level: 
X-Spam-Status: No, score=-1.997 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, HTTPS_HTTP_MISMATCH=0.1, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=unavailable autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id q6MduqiNHlWz for <quic@ietfa.amsl.com>; Fri, 19 Jun 2020 17:56:16 -0700 (PDT)
Received: from mail-io1-xd32.google.com (mail-io1-xd32.google.com [IPv6:2607:f8b0:4864:20::d32]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id DAA9B3A0F7A for <quic@ietf.org>; Fri, 19 Jun 2020 17:56:15 -0700 (PDT)
Received: by mail-io1-xd32.google.com with SMTP id r2so13449465ioo.4 for <quic@ietf.org>; Fri, 19 Jun 2020 17:56:15 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=yX21H14U50qkycTl3XSpc/GVu5iBjTrehIS2mRVB3jk=; b=XoIaqkUg4gEbMEmwKzGuk3b6Uq12M2XeuZ8RZydaJhDj/hbPbYv54zSwImwfivwHyY uwJ2TRdwBP4zA1/Ab0GOxS0tvnIhXd2d0um3PEY+bxg4a/G8wFxvURU6yOlxkgux/0Gf 9BRlqRP/6qX19e/+WIKyAYqDDdb4IUZriQkNbjYx8xZZDl/GXoq/PbhtmelCgplsp9L+ 1IJ2rP7vnN/mVDs6zEwxVqGL2TogiOsv8InVQfWllB66kMi15T82GG3CNX65PP17Gqo8 S2VoC/GkZMF78e1E4d2msurOHtrszcDFREwcnvzff+tyUdICtfImG5kaQflrn+iAlxLq yrsw==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=yX21H14U50qkycTl3XSpc/GVu5iBjTrehIS2mRVB3jk=; b=Xe0zDrF9Dvyu2hQYho77EIlxymNj0N3lki2/e9D7RTtNXHpbMn32XM5NXKLYPYXPQr kT6IgJl3UF5xBkf5cHpSUoHGPDjE8jAHImuCQmWRptWj7TSDAvqcr782z4RbRTQs8tsB l93AKasGu6oZuUu9MUCh+z8Y59XZke8uvOswJK0mXQHIi9ObdIWxbuWTuHmta717iYlC wpGWwX7ctGXXq7FkjjulUEVi8a13jDwEsJSR6659SJ8DZs6vb+NYC8hsUKoyEZTjBb12 vTmlcWy5J0YveQfufGBZC+GcqfU3UgBNwWUiAes1YFstRuoeVeZnu6ilqXPczd8nVcA0 Y/Rw==
X-Gm-Message-State: AOAM530X3JU6cndTEB6BUH3J7cGb6q1nY4ZOXJ8R6JwxncxyCvKhiKjS PVG+XzBVre5zxKMkt9vHDX1Cj+W6Dce0ellhG5I=
X-Google-Smtp-Source: ABdhPJz61GG1jYzAyjdRpMHFyYzGU27AD5r8+aap9lQ7TPtXmcGqG+BGCtb3dm/RqiFktis9ldIBAM+keH74/EKSHro=
X-Received: by 2002:a5d:9a13:: with SMTP id s19mr7158330iol.20.1592614574943;  Fri, 19 Jun 2020 17:56:14 -0700 (PDT)
MIME-Version: 1.0
References: <CAM4esxQBqfrz24riPQA_VGKcGp_TzW0pqb97KfFMtNdW9pUfDg@mail.gmail.com> <833A693C-62E6-4889-9954-FCE65A839A7C@eggert.org> <CAKcm_gPMO2DtqvKucqVw0zDjSniSOmFD4p1Tp4YLjr9WSWdEUw@mail.gmail.com> <CAJU8_nUN42gGmQof24XD9-EjXedyzcarDyRP8MGe1qW-BZ=+Aw@mail.gmail.com> <9cd91c24-c730-22a4-7aa0-baf61613b3ce@huitema.net> <f4922cdb59014202900de44cc5fea0ff@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQvwkTvpUcu6-+W5zWo22m-R1jvN7DcCpXfuw8Hb55qsw@mail.gmail.com> <95dd02c92b32472d9cab0dd47b98c637@usma1ex-dag1mb5.msg.corp.akamai.com> <CAM4esxQxxXn27rZEY75-AsHD5VF0fqiV1VDyeSrzQ=-sM7JNCg@mail.gmail.com> <9c2e300c30f74d1794d11cf4334ea07b@usma1ex-dag1mb5.msg.corp.akamai.com> <2c40f3d9-fa40-9834-ac30-36bc9a1a6303@huitema.net> <CA+9kkMBQt001xOVgT=9G8YOOTOJ+9S=OWDwGEeYuKVm46Cq3iQ@mail.gmail.com> <3bb42dfc-17ba-c5c8-03f7-35428756b4c2@huitema.net> <CAM4esxRWVRjVhxyYuuzwDGq_wfTjQHkY6KHG2rEPErO2aHXA0w@mail.gmail.com> <f9e2c611-bb4d-bc80-dfe3-e323a08bfc5b@huitema.net>
In-Reply-To: <f9e2c611-bb4d-bc80-dfe3-e323a08bfc5b@huitema.net>
From: Martin Duke <martin.h.duke@gmail.com>
Date: Fri, 19 Jun 2020 17:56:03 -0700
Message-ID: <CAM4esxQsZba_Ck80hNX23Yd6tZfgLTfG-VNWvBDhn9V7UfNnbg@mail.gmail.com>
Subject: Re: Is the invariants draft really standards track?
To: Christian Huitema <huitema@huitema.net>
Cc: Ted Hardie <ted.ietf@gmail.com>, Jared Mauch <jared@puck.nether.net>,  Kyle Rose <krose@krose.org>, "Lubashev, Igor" <ilubashe=40akamai.com@dmarc.ietf.org>,  Ian Swett <ianswett=40google.com@dmarc.ietf.org>, Lars Eggert <lars@eggert.org>, IETF QUIC WG <quic@ietf.org>
Content-Type: multipart/alternative; boundary="00000000000000f2dc05a8797ac9"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/zuMIpP3tTqk-pFpCzyot76O0Isc>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Sat, 20 Jun 2020 00:56:19 -0000

--00000000000000f2dc05a8797ac9
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Ah, I understand now. Thanks for explaining it again.

Oddly enough, the load-balancing part of QUIC-LB provides some DoS
protection by making it easier to check for valid conn IDs on 1RTT packets
further away from the server.

On Fri, Jun 19, 2020 at 4:52 PM Christian Huitema <huitema@huitema.net>
wrote:

>
> On 6/19/2020 4:00 PM, Martin Duke wrote:
>
> >  The problem of the DOS protection box is to drop as much traffic as
> possible while still letting some good in, and also without causing
> performance hits when there are no attacks. There are indeed some thorny
> traffic management issues there. More information helps, and Ben's
> suggestion to look at DOTS is certainly on point.
>
> I am not sure how any of this can work when inbound 1RTT packets can come
> with essentially random IP/port and CID and potentially be valid, given
> there might have been a migration. Some of the QUIC-LB modes at least mak=
e
> it hard to guess your way to a valid CID.
>
> The DOS box does not have to worry about what kind of traffic is coming
> in. It just has to open a context for the 5 tuple, and check whether it
> sees 1-RTT packets coming back. And then maybe count the volume of 1-RTT
> packets coming back.
>
> The worry is that one of the bots might start a legitimate connection,
> then disclose its five tuple to the rest of the botnet. The whole botnet
> can then spoof the 5 tuple that was just pin-holed by the DOS box. A simp=
le
> "open-close" logic is thus not good enough. The DOS box must also enforce
> some kind of rate limiting per 5 tuple.
>
> Which also means that if a botnet can predict the 5 tuple used by a
> legitimate connection and then spoof it, it can DOS it. Once you start
> digging that particular rabbit hole, the joy never stops...
>
> -- Christian Huitema
>
>
>
> DOTS is an intriguing idea, but again I'm not sure what signatures one
> could use to configure it.
>
> On Fri, Jun 19, 2020 at 3:43 PM Christian Huitema <huitema@huitema.net>
> wrote:
>
>>
>> On 6/19/2020 3:05 PM, Ted Hardie wrote:
>>
>> Hi Christian,
>>
>> A question in-line,
>>
>> On Fri, Jun 19, 2020 at 2:49 PM Christian Huitema <huitema@huitema.net>
>> wrote:
>>
>>> When under DOS attack, you want to "minimize blowback", i.e., as much a=
s
>>> possible avoid generating packets in response to attack traffic. So, ye=
s, a
>>> server may choose to not send stateless resets to anyone when under att=
ack;
>>> in fact, my recommendation would be that a server SHOULD NOT send state=
less
>>> resets to anyone when under attack.
>>>
>>> That said, Igor raised an interesting point about return traffic. It
>>> would be very nice if DOS protection boxes could distinguish between
>>> "validated traffic" that the server presumably intends to process, and
>>> "unsolicited traffic" that will just consume resource. The box can then
>>> reserve some share of the resource for validated traffic, and place the
>>> rest of the traffic in a lower priority queue. Fine, but there needs to=
 be
>>> a test. The classic test is that incoming traffic is "validated" if the
>>> protection box can match it with return traffic coming from the server =
--
>>> for some definition of matching.
>>>
>>> From that point of view, stateless reset is definitely not helpful. But
>>> problematic traffic goes beyond that. The server will reply to a client=
's
>>> initial with a server's initial packet. Does that validate the response
>>> traffic? OK, maybe the protection box can programmed to only validate
>>> traffic if it sees 1RTT packets. But many servers will send 1-RTT packe=
ts
>>> as 0.5 RTT. Does that validate the response traffic?
>>>
>>> We might say that traffic is validated when the handshake is confirmed,
>>> but the protection box does not understand the TLS handshake, it just s=
ees
>>> packet types and packet sizes. It cannot distinguish between 0.5RTT dat=
a
>>> and 1RTT data, and thus the closest approximation of "validation" would=
 be
>>> seeing more than an initial window worth of traffic coming back from th=
e
>>> server. That does not sound great.
>>>
>>> On the other hand, things get much better if the server under attack ca=
n
>>> adopt a defensive posture and help the DOS protection box do its job.
>>> Suppose that a server can detect that it is under attack -- or be
>>> explicitly configured so. The simplest defensive posture would be to (1=
)
>>> disable stateless reset and (2) not send any 0.5RTT packet, including i=
n
>>> response to 0-RTT. The protection boxes can at that point take the 1-RT=
T
>>> packets from the server as indicating validation.
>>>
>> Perhaps I'm misreading you here, but this sounds like you expect the
>> protection boxes to be able to distinguish when servers see themselves a=
s
>> under DOS from when they don't, so that they can tell that the lack of
>> 0.5RTT is an indication of an attack response.  Given distribution patte=
rns
>> of DOS attacks, I'm struggling to see whether that will commonly be the
>> case.
>>
>> Maybe I wasn't too clear. I don't think that DOS protection boxes are
>> trying to take directions from servers. I also don't believe that DOS
>> protection boxes have any practical means of distinguishing between 0.5 =
RTT
>> and 1 RTT traffic. That might actually be a problem.
>>
>>
>> You could, of course, always put handshake traffic into low-priority
>> queues until you see the 1-RTT packets that validate the server's
>> interest.  That would make 1-RTT traffic effectively a path signal of
>> validation.
>>
>> Yes, that's more or less what I was considering. But of course it only
>> works if the server refrains from sending 1RTT packets until the handsha=
ke
>> is confirmed. Otherwise, the protection boxes will have to count the num=
ber
>> of 1RTT packets coming back from the server, or maybe the amount of data
>> coming from the server, and only consider the traffic validated if numbe=
r
>> of packets and amount of data are larger than common values of IW. You
>> might end up with 3 classes instead of 2 -- validated if 1RTT > IW,
>> validation in progress if 1RTT see but < IW, not validated yet if no 1RT=
T
>> seen.
>>
>> My concern with that is that using those low priority queues during the
>> handshake phase seems likely to result in worse latency and increased ri=
sk
>> of packet loss (which can be tricky during that phase).  That seems a he=
avy
>> price to pay during non-attack times for better protection when under
>> attack.
>>
>> The problem of the DOS protection box is to drop as much traffic as
>> possible while still letting some good in, and also without causing
>> performance hits when there are no attacks. There are indeed some thorny
>> traffic management issues there. More information helps, and Ben's
>> suggestion to look at DOTS is certainly on point.
>>
>> I am just doing a thought experiment so far. One immediate observation i=
s
>> that in the absence of other data, DOS protection designers are likely t=
o
>> look at packet types, and certainly reason about 1RTT packets versus lon=
g
>> header packets. This is definitely an ossification risk.
>>
>> The other observation is that we have to be careful about the realism of
>> thought experiments. If I remember correctly, the payload of the Blaster
>> worm was something like:
>>
>>     On each bot
>>
>>         Open 256 threads
>>
>>             In each thread, loop on "GET some large page from the server=
"
>>
>> Reasoning about validated traffic would not stop that. But reasoning
>> about validated traffic forces the attacker to disclose the "real" IP
>> addresses of the attacking bots, which then enables a second line of
>> defense.
>>
>> -- Christian Huitema
>>
>>
>>
>>
>> Am I misunderstanding how this is distinguished?
>>
>> Clue appreciated,
>>
>> Ted
>>
>>
>>
>>> Maybe we should specify that.
>>>
>>> -- Christian Huitema
>>> On 6/19/2020 1:42 PM, Lubashev, Igor wrote:
>>>
>>> > There is no need for servers to decrypt CIDs in QUIC-LB. Presumably
>>> the server has a lookup table for its CIDs.
>>>
>>>
>>>
>>> Sending a stateless reset in response to a junk packet would cost more
>>> CPU than verifying CID integrity.  But, yes, a server may choose to not
>>> send stateless resets to anyone when under attack.
>>>
>>>
>>>
>>>
>>>
>>> *From:* Martin Duke <martin.h.duke@gmail.com> <martin.h.duke@gmail.com>
>>> *Sent:* Friday, June 19, 2020 2:44 PM
>>>
>>> >Unfortunately, Retry system protects only server's memory state and
>>> some CPU cycles spent on crypto.  (Servers still need to decrypt CID to
>>> decide it is invalid, and if the attacker is clever enough to establish=
 one
>>> valid connection and use that CID in a flood, the server will also be
>>> decrypting packets.)
>>>
>>>
>>>
>>> There is no need for servers to decrypt CIDs in QUIC-LB. Presumably the
>>> server has a lookup table for its CIDs.
>>>
>>>
>>>
>>> It is true that Retry Services (and indeed, the Retry concept as a
>>> whole) does nothing to protect network capacity.
>>>
>>>
>>>
>>> On Fri, Jun 19, 2020 at 8:08 AM Lubashev, Igor <ilubashe@akamai.com>
>>> wrote:
>>>
>>> It looks like
>>> https://tools.ietf.org/html/draft-ietf-quic-load-balancers-02#section-5
>>> <https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__tools.ietf.org_=
html_draft-2Dietf-2Dquic-2Dload-2Dbalancers-2D02-23section-2D5&d=3DDwMFaQ&c=
=3D96ZbZZcaMF4w0F4jpN6LZg&r=3DDjn3bQ5uNJDPM_2skfL3rW1tzcIxyjUZdn_m55KPmlo&m=
=3DeTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTKljvj-ys&s=3D1x7AN2a51X_zV5xSyK0uCL8v=
Z5cugD3n4lbFWZDqVQg&e=3D>
>>> is an excellent discussion of Retry mechanics.  It definitely deserves =
a
>>> reference from this manageability draft.
>>>
>>>
>>>
>>> The Retry mechanisms described in LB draft are all cooperating boxes,
>>> and servers must be aware of them.  Unfortunately, Retry system protect=
s
>>> only server's memory state and some CPU cycles spent on crypto.  (Serve=
rs
>>> still need to decrypt CID to decide it is invalid, and if the attacker =
is
>>> clever enough to establish one valid connection and use that CID in a
>>> flood, the server will also be decrypting packets.)  Retry does nothing=
 to
>>> protect network resources.
>>>
>>>
>>>
>>> The PR I opened (https://github.com/quicwg/ops-drafts/pull/94
>>> <https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__github.com_quic=
wg_ops-2Ddrafts_pull_94&d=3DDwMFaQ&c=3D96ZbZZcaMF4w0F4jpN6LZg&r=3DDjn3bQ5uN=
JDPM_2skfL3rW1tzcIxyjUZdn_m55KPmlo&m=3DeTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTK=
ljvj-ys&s=3DVpvhQ9n79rANIk3O5Sm7PZyToFQEennoPt9iqFpWbq8&e=3D>)
>>> is about uncooperating devices that try to mitigate volumetric network
>>> attacks.
>>>
>>>
>>>
>>>
>>>
>>> *From:* Martin Duke <martin.h.duke@gmail.com>
>>> *Sent:* Wednesday, June 17, 2020 8:16 PM
>>>
>>> Hi Igor, you might want to check out the "Retry Services" bit of the
>>> QUIC-LB draft. This has something to do with the DDoS use case you disc=
uss.
>>>
>>>
>>>
>>> On Wed, May 27, 2020 at 9:07 AM Lubashev, Igor <ilubashe@akamai.com>
>>> wrote:
>>>
>>> I=E2=80=99m working on a manageability draft PR for this (how to rate l=
imit UDP
>>> to reduce disruption to QUIC if you have to rate limit UDP).  ETA end o=
f
>>> the week (if I do not get pulled into something again).
>>>
>>>
>>>
>>> The relevant observation is that DDoS with UDP that is indistinguishabl=
e
>>> from QUIC will happen.  UDP is already the most prevalent DDoS vector,
>>> since it is easy for a compromised non-admin app to send a flood of hug=
e
>>> UDP packets (with TCP you get throttled by the congestion controller). =
 So
>>> there WILL be DDoS protection devices out there to try to mitigate the
>>> problem, possibly by observing both directions of the flow and deciding
>>> whether a packet belongs to a valid flow or not.
>>>
>>>
>>>
>>> Since such middle boxes will be created, the more explicit and normativ=
e
>>> Invariants are about what one can expect, the less such middle boxes ma=
y
>>> decide for themselves.  For example (I did not think long about it), if
>>> some elements of path validation could land into Invariants (roughly, =
=E2=80=9Cno
>>> more than X packets/bytes can be sent on a new path w/o a return packet=
=E2=80=9D),
>>> a DDoS middle box may use this fact and active connection migration mig=
ht
>>> still have a chance during an attack (NAT rebinding could be linked by =
DDoS
>>> boxes to an old connection via unchanged CID).
>>>
>>>
>>>
>>>    - Igor
>>>
>>>
>>>
>>>
>>>
>>> *From:* Christian Huitema <huitema@huitema.net>
>>> *Sent:* Wednesday, May 27, 2020 11:34 AM
>>>
>>> On 5/27/2020 8:28 AM, Kyle Rose wrote:
>>>
>>> On Wed, May 27, 2020 at 10:34 AM Ian Swett <ianswett=3D
>>> 40google.com@dmarc.ietf.org> wrote:
>>>
>>> I was agreeing with MT, but I'm happy to see some more MUSTs added if
>>> people feel that'd be helpful.
>>>
>>>
>>>
>>> Coincidentally, we were just talking about this internally at Akamai
>>> yesterday. IMO, an invariants document isn't really helpful if it isn't
>>> normative, and for it to be normative it (or a related practices doc fo=
r
>>> operators) really needs to spell out clear boundaries for operators wit=
h
>>> MUSTs..
>>>
>>>
>>>
>>> The example that came up yesterday was around operators filtering QUIC
>>> in the event of a DDoS: one recommendation based on some conversations
>>> going back at least to Prague 2019 was to hash packets on 4-tuple and
>>> filter those below a hash value chosen for a desired ingress limit inst=
ead
>>> of doing what most operators do with UDP today, which is to cap UDP
>>> throughput and just drop packets randomly or tail drop.
>>>
>>> Interesting. Did they consider using the CID, or a fraction of it? This
>>> looks entirely like the scenario for which we developed stateless retry=
.
>>>
>>>
>>>
>>> This recommendation certainly imposes some constraints on future
>>> protocol development that motivate new invariants: for instance, it wou=
ld
>>> preclude sharding a connection across multiple source ports (not that t=
here
>>> is necessarily a reason to do this; it's just an example). But more
>>> importantly, it goes beyond invariants: it's one among many practices
>>> compatible with the current set of invariants, some reasonable and some
>>> terrible.
>>>
>>> This would break the "preferred address" redirection. Preferred address
>>> migration may or may not be spelled out in the invariants.
>>>
>>>
>>>
>>> Operators are going to do things to QUIC traffic, so it would be good t=
o
>>> offer them recommendations that are compatible with broad deployability=
.
>>>
>>>
>>>
>>> Yes, we do need the invariants for that.
>>>
>>> -- Christian Huitema
>>>
>>>

--00000000000000f2dc05a8797ac9
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Ah, I understand now. Thanks for explaining it again.<div>=
<br></div><div>Oddly enough, the load-balancing part of QUIC-LB provides so=
me DoS protection by making it easier to check for valid conn IDs on 1RTT p=
ackets further away from the server.</div></div><br><div class=3D"gmail_quo=
te"><div dir=3D"ltr" class=3D"gmail_attr">On Fri, Jun 19, 2020 at 4:52 PM C=
hristian Huitema &lt;<a href=3D"mailto:huitema@huitema.net">huitema@huitema=
.net</a>&gt; wrote:<br></div><blockquote class=3D"gmail_quote" style=3D"mar=
gin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1=
ex">
 =20
   =20
 =20
  <div>
    <p><br>
    </p>
    <div>On 6/19/2020 4:00 PM, Martin Duke
      wrote:<br>
    </div>
    <blockquote type=3D"cite">
     =20
      <div dir=3D"ltr">&gt;=C2=A0
        The problem of the DOS protection box is to drop as much traffic
        as possible while still letting some good in, and also without
        causing performance hits when there are no attacks. There are
        indeed some thorny traffic management issues there. More
        information helps, and Ben&#39;s suggestion to look at DOTS is
        certainly on point.
        <div><br>
        </div>
        <div>I am not sure how any of this can work when inbound=C2=A01RTT
          packets can come with essentially random IP/port and CID and
          potentially be valid, given there might have been a migration.
          Some of the QUIC-LB modes at least make it hard to guess your
          way to a valid=C2=A0CID.</div>
      </div>
    </blockquote>
    <p>The DOS box does not have to worry about what kind of traffic is
      coming in. It just has to open a context for the 5 tuple, and
      check whether it sees 1-RTT packets coming back. And then maybe
      count the volume of 1-RTT packets coming back.</p>
    <p>The worry is that one of the bots might start a legitimate
      connection, then disclose its five tuple to the rest of the
      botnet. The whole botnet can then spoof the 5 tuple that was just
      pin-holed by the DOS box. A simple &quot;open-close&quot; logic is th=
us not
      good enough. The DOS box must also enforce some kind of rate
      limiting per 5 tuple. <br>
    </p>
    <p>Which also means that if a botnet can predict the 5 tuple used by
      a legitimate connection and then spoof it, it can DOS it. Once you
      start digging that particular rabbit hole, the joy never stops...</p>
    <p>-- Christian Huitema<br>
    </p>
    <p><br>
    </p>
    <blockquote type=3D"cite">
      <div dir=3D"ltr">
        <div><br>
        </div>
        <div>DOTS is an intriguing idea, but again I&#39;m not sure what
          signatures one could use to configure it.</div>
      </div>
      <br>
      <div class=3D"gmail_quote">
        <div dir=3D"ltr" class=3D"gmail_attr">On Fri, Jun 19, 2020 at 3:43
          PM Christian Huitema &lt;<a href=3D"mailto:huitema@huitema.net" t=
arget=3D"_blank">huitema@huitema.net</a>&gt; wrote:<br>
        </div>
        <blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex=
;border-left:1px solid rgb(204,204,204);padding-left:1ex">
          <div>
            <p><br>
            </p>
            <div>On 6/19/2020 3:05 PM, Ted Hardie wrote:<br>
            </div>
            <blockquote type=3D"cite">
              <div dir=3D"ltr">
                <div dir=3D"ltr">
                  <div>Hi Christian,</div>
                  <div><br>
                  </div>
                  <div>A question in-line,<br>
                  </div>
                </div>
                <br>
                <div class=3D"gmail_quote">
                  <div dir=3D"ltr" class=3D"gmail_attr">On Fri, Jun 19, 202=
0
                    at 2:49 PM Christian Huitema &lt;<a href=3D"mailto:huit=
ema@huitema.net" target=3D"_blank">huitema@huitema.net</a>&gt;
                    wrote:<br>
                  </div>
                  <blockquote class=3D"gmail_quote" style=3D"margin:0px 0px=
 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
                    <div>
                      <p>When under DOS attack, you want to &quot;minimize
                        blowback&quot;, i.e., as much as possible avoid
                        generating packets in response to attack
                        traffic. So, yes, a server may choose to not
                        send stateless resets to anyone when under
                        attack; in fact, my recommendation would be that
                        a server SHOULD NOT send stateless resets to
                        anyone when under attack.</p>
                      <p>That said, Igor raised an interesting point
                        about return traffic. It would be very nice if
                        DOS protection boxes could distinguish between
                        &quot;validated traffic&quot; that the server presu=
mably
                        intends to process, and &quot;unsolicited traffic&q=
uot;
                        that will just consume resource. The box can
                        then reserve some share of the resource for
                        validated traffic, and place the rest of the
                        traffic in a lower priority queue. Fine, but
                        there needs to be a test. The classic test is
                        that incoming traffic is &quot;validated&quot; if t=
he
                        protection box can match it with return traffic
                        coming from the server -- for some definition of
                        matching.</p>
                      <p>From that point of view, stateless reset is
                        definitely not helpful. But problematic traffic
                        goes beyond that. The server will reply to a
                        client&#39;s initial with a server&#39;s initial pa=
cket.
                        Does that validate the response traffic? OK,
                        maybe the protection box can programmed to only
                        validate traffic if it sees 1RTT packets. But
                        many servers will send 1-RTT packets as 0.5 RTT.
                        Does that validate the response traffic?</p>
                      <p>We might say that traffic is validated when the
                        handshake is confirmed, but the protection box
                        does not understand the TLS handshake, it just
                        sees packet types and packet sizes. It cannot
                        distinguish between 0.5RTT data and 1RTT data,
                        and thus the closest approximation of
                        &quot;validation&quot; would be seeing more than an
                        initial window worth of traffic coming back from
                        the server. That does not sound great.</p>
                      <p>On the other hand, things get much better if
                        the server under attack can adopt a defensive
                        posture and help the DOS protection box do its
                        job. Suppose that a server can detect that it is
                        under attack -- or be explicitly configured so.
                        The simplest defensive posture would be to (1)
                        disable stateless reset and (2) not send any
                        0.5RTT packet, including in response to 0-RTT.
                        The protection boxes can at that point take the
                        1-RTT packets from the server as indicating
                        validation.</p>
                    </div>
                  </blockquote>
                  <div>Perhaps I&#39;m misreading you here, but this sounds
                    like you expect the protection boxes to be able to
                    distinguish when servers see themselves as under DOS
                    from when they don&#39;t, so that they can tell that th=
e
                    lack of 0.5RTT is an indication of an attack
                    response.=C2=A0 Given distribution patterns of DOS
                    attacks, I&#39;m struggling to see whether that will
                    commonly be the case.</div>
                </div>
              </div>
            </blockquote>
            <p>Maybe I wasn&#39;t too clear. I don&#39;t think that DOS
              protection boxes are trying to take directions from
              servers. I also don&#39;t believe that DOS protection boxes
              have any practical means of distinguishing between 0.5 RTT
              and 1 RTT traffic. That might actually be a problem.<br>
            </p>
            <blockquote type=3D"cite">
              <div dir=3D"ltr">
                <div class=3D"gmail_quote">
                  <div><br>
                  </div>
                  <div>You could, of course, always put handshake
                    traffic into low-priority queues until you see the
                    1-RTT packets that validate the server&#39;s interest.=
=C2=A0
                    That would make 1-RTT traffic effectively a path
                    signal of validation.=C2=A0 </div>
                </div>
              </div>
            </blockquote>
            <p>Yes, that&#39;s more or less what I was considering. But of
              course it only works if the server refrains from sending
              1RTT packets until the handshake is confirmed. Otherwise,
              the protection boxes will have to count the number of 1RTT
              packets coming back from the server, or maybe the amount
              of data coming from the server, and only consider the
              traffic validated if number of packets and amount of data
              are larger than common values of IW. You might end up with
              3 classes instead of 2 -- validated if 1RTT &gt; IW,
              validation in progress if 1RTT see but &lt; IW, not
              validated yet if no 1RTT seen.<br>
            </p>
            <blockquote type=3D"cite">
              <div dir=3D"ltr">
                <div class=3D"gmail_quote">
                  <div>My concern with that is that using those low
                    priority queues during the handshake phase seems
                    likely to result in worse latency and increased risk
                    of packet loss (which can be tricky during that
                    phase).=C2=A0 That seems a heavy price to pay during
                    non-attack times for better protection when under
                    attack.</div>
                </div>
              </div>
            </blockquote>
            <p>The problem of the DOS protection box is to drop as much
              traffic as possible while still letting some good in, and
              also without causing performance hits when there are no
              attacks. There are indeed some thorny traffic management
              issues there. More information helps, and Ben&#39;s suggestio=
n
              to look at DOTS is certainly on point.</p>
            <p>I am just doing a thought experiment so far. One
              immediate observation is that in the absence of other
              data, DOS protection designers are likely to look at
              packet types, and certainly reason about 1RTT packets
              versus long header packets. This is definitely an
              ossification risk.</p>
            <p>The other observation is that we have to be careful about
              the realism of thought experiments. If I remember
              correctly, the payload of the Blaster worm was something
              like:</p>
            <p>=C2=A0=C2=A0=C2=A0 On each bot</p>
            <p>=C2=A0=C2=A0=C2=A0 =C2=A0=C2=A0=C2=A0 Open 256 threads</p>
            <p>=C2=A0=C2=A0=C2=A0 =C2=A0=C2=A0=C2=A0 =C2=A0=C2=A0=C2=A0 In =
each thread, loop on &quot;GET some large page
              from the server&quot;</p>
            <p>Reasoning about validated traffic would not stop that.
              But reasoning about validated traffic forces the attacker
              to disclose the &quot;real&quot; IP addresses of the attackin=
g bots,
              which then enables a second line of defense.<br>
            </p>
            <p>-- Christian Huitema<br>
            </p>
            <p><br>
            </p>
            <p><br>
            </p>
            <blockquote type=3D"cite">
              <div dir=3D"ltr">
                <div class=3D"gmail_quote">
                  <div><br>
                  </div>
                  <div>Am I misunderstanding how this is distinguished?</di=
v>
                  <div><br>
                  </div>
                  <div>Clue appreciated,</div>
                  <div><br>
                  </div>
                  <div>Ted<br>
                  </div>
                  <div><br>
                  </div>
                  <div>=C2=A0</div>
                  <blockquote class=3D"gmail_quote" style=3D"margin:0px 0px=
 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
                    <div>
                      <p>Maybe we should specify that.</p>
                      <p>-- Christian Huitema<br>
                      </p>
                      <div>On 6/19/2020 1:42 PM, Lubashev, Igor wrote:<br>
                      </div>
                      <blockquote type=3D"cite">
                        <div>
                          <p class=3D"MsoNormal">&gt; There is no need for
                            servers to decrypt CIDs in QUIC-LB.
                            Presumably the server has a lookup table for
                            its CIDs.</p>
                          <p class=3D"MsoNormal">=C2=A0</p>
                          <p class=3D"MsoNormal">Sending a stateless reset
                            in response to a junk packet would cost more
                            CPU than verifying CID integrity.=C2=A0 But, ye=
s,
                            a server may choose to not send stateless
                            resets to anyone when under attack.</p>
                          <p class=3D"MsoNormal">=C2=A0</p>
                          <p class=3D"MsoNormal">=C2=A0</p>
                          <div style=3D"border-color:currentcolor currentco=
lor currentcolor blue;border-style:none none none solid;border-width:medium=
 medium medium 1.5pt;padding:0in 0in 0in 4pt">
                            <div>
                              <div style=3D"border-color:rgb(225,225,225) c=
urrentcolor currentcolor;border-style:solid none none;border-width:1pt medi=
um medium;padding:3pt 0in 0in">
                                <p class=3D"MsoNormal"><b>From:</b> Martin
                                  Duke <a href=3D"mailto:martin.h.duke@gmai=
l.com" target=3D"_blank">&lt;martin.h.duke@gmail.com&gt;</a>
                                  <br>
                                  <b>Sent:</b> Friday, June 19, 2020
                                  2:44 PM<br>
                                  <br>
                                </p>
                              </div>
                            </div>
                            <div>
                              <p class=3D"MsoNormal">&gt;Unfortunately,
                                Retry system protects only server&#39;s
                                memory state and some CPU cycles spent
                                on crypto.=C2=A0 (Servers still need to
                                decrypt CID to decide it is invalid, and
                                if the attacker is clever enough to
                                establish one valid connection and use
                                that CID in a flood, the server will
                                also be decrypting packets.)=C2=A0=C2=A0</p=
>
                              <div>
                                <p class=3D"MsoNormal">=C2=A0</p>
                              </div>
                              <div>
                                <p class=3D"MsoNormal">There is no need
                                  for servers to decrypt CIDs in
                                  QUIC-LB. Presumably the server has a
                                  lookup table for its CIDs.</p>
                              </div>
                              <div>
                                <p class=3D"MsoNormal">=C2=A0</p>
                              </div>
                              <div>
                                <p class=3D"MsoNormal">It is true that
                                  Retry Services (and indeed, the Retry
                                  concept as a whole) does nothing to
                                  protect network capacity.</p>
                              </div>
                            </div>
                            <p class=3D"MsoNormal">=C2=A0</p>
                            <div>
                              <div>
                                <p class=3D"MsoNormal">On Fri, Jun 19,
                                  2020 at 8:08 AM Lubashev, Igor &lt;<a hre=
f=3D"mailto:ilubashe@akamai.com" target=3D"_blank">ilubashe@akamai.com</a>&=
gt;
                                  wrote:</p>
                              </div>
                              <blockquote style=3D"border-color:currentcolo=
r currentcolor currentcolor rgb(204,204,204);border-style:none none none so=
lid;border-width:medium medium medium 1pt;padding:0in 0in 0in 6pt;margin-le=
ft:4.8pt;margin-right:0in">
                                <div>
                                  <div>
                                    <p class=3D"MsoNormal">It looks like <a=
 href=3D"https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__tools.ietf.=
org_html_draft-2Dietf-2Dquic-2Dload-2Dbalancers-2D02-23section-2D5&amp;d=3D=
DwMFaQ&amp;c=3D96ZbZZcaMF4w0F4jpN6LZg&amp;r=3DDjn3bQ5uNJDPM_2skfL3rW1tzcIxy=
jUZdn_m55KPmlo&amp;m=3DeTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTKljvj-ys&amp;s=3D=
1x7AN2a51X_zV5xSyK0uCL8vZ5cugD3n4lbFWZDqVQg&amp;e=3D" target=3D"_blank">
https://tools.ietf.org/html/draft-ietf-quic-load-balancers-02#section-5</a>
                                      is an excellent discussion of
                                      Retry mechanics.=C2=A0 It definitely
                                      deserves a reference from this
                                      manageability draft.</p>
                                    <p class=3D"MsoNormal">=C2=A0</p>
                                    <p class=3D"MsoNormal">The Retry
                                      mechanisms described in LB draft
                                      are all cooperating boxes, and
                                      servers must be aware of them.=C2=A0
                                      Unfortunately, Retry system
                                      protects only server&#39;s memory
                                      state and some CPU cycles spent on
                                      crypto.=C2=A0 (Servers still need to
                                      decrypt CID to decide it is
                                      invalid, and if the attacker is
                                      clever enough to establish one
                                      valid connection and use that CID
                                      in a flood, the server will also
                                      be decrypting packets.)=C2=A0 Retry
                                      does nothing to protect network
                                      resources.</p>
                                    <p class=3D"MsoNormal">=C2=A0</p>
                                    <p class=3D"MsoNormal">The PR I opened
                                      (<a href=3D"https://urldefense.proofp=
oint.com/v2/url?u=3Dhttps-3A__github.com_quicwg_ops-2Ddrafts_pull_94&amp;d=
=3DDwMFaQ&amp;c=3D96ZbZZcaMF4w0F4jpN6LZg&amp;r=3DDjn3bQ5uNJDPM_2skfL3rW1tzc=
IxyjUZdn_m55KPmlo&amp;m=3DeTT-BoZ1fMitywKRVSxpU3js0lhO0qkspYTKljvj-ys&amp;s=
=3DVpvhQ9n79rANIk3O5Sm7PZyToFQEennoPt9iqFpWbq8&amp;e=3D" target=3D"_blank">=
https://github.com/quicwg/ops-drafts/pull/94</a>)
                                      is about uncooperating devices
                                      that try to mitigate volumetric
                                      network attacks.</p>
                                    <p class=3D"MsoNormal">=C2=A0</p>
                                    <p class=3D"MsoNormal">=C2=A0</p>
                                    <div style=3D"border-color:currentcolor=
 currentcolor currentcolor blue;border-style:none none none solid;border-wi=
dth:medium medium medium 1.5pt;padding:0in 0in 0in 4pt">
                                      <div>
                                        <div style=3D"border-color:rgb(225,=
225,225) currentcolor currentcolor;border-style:solid none none;border-widt=
h:1pt medium medium;padding:3pt 0in 0in">
                                          <p class=3D"MsoNormal" style=3D"m=
argin-bottom:12pt"><b>From:</b>
                                            Martin Duke &lt;<a href=3D"mail=
to:martin.h.duke@gmail.com" target=3D"_blank">martin.h.duke@gmail.com</a>&g=
t;
                                            <br>
                                            <b>Sent:</b> Wednesday, June
                                            17, 2020 8:16 PM</p>
                                        </div>
                                      </div>
                                      <div>
                                        <p class=3D"MsoNormal">Hi Igor,
                                          you might want to check out
                                          the &quot;Retry Services&quot; bi=
t of
                                          the QUIC-LB draft. This has
                                          something to do with the DDoS
                                          use case you discuss.</p>
                                      </div>
                                      <p class=3D"MsoNormal">=C2=A0</p>
                                      <div>
                                        <div>
                                          <p class=3D"MsoNormal">On Wed,
                                            May 27, 2020 at 9:07 AM
                                            Lubashev, Igor &lt;<a href=3D"m=
ailto:ilubashe@akamai.com" target=3D"_blank">ilubashe@akamai.com</a>&gt;
                                            wrote:</p>
                                        </div>
                                        <blockquote style=3D"border-color:c=
urrentcolor currentcolor currentcolor rgb(204,204,204);border-style:none no=
ne none solid;border-width:medium medium medium 1pt;padding:0in 0in 0in 6pt=
;margin:5pt 0in 5pt 4.8pt">
                                          <div>
                                            <div>
                                              <p class=3D"MsoNormal">I=E2=
=80=99m
                                                working on a
                                                manageability draft PR
                                                for this (how to rate
                                                limit UDP to reduce
                                                disruption to QUIC if
                                                you have to rate limit
                                                UDP).=C2=A0 ETA end of the
                                                week (if I do not get
                                                pulled into something
                                                again).</p>
                                              <p class=3D"MsoNormal">=C2=A0=
</p>
                                              <p class=3D"MsoNormal">The
                                                relevant observation is
                                                that DDoS with UDP that
                                                is indistinguishable
                                                from QUIC will happen.=C2=
=A0
                                                UDP is already the most
                                                prevalent DDoS vector,
                                                since it is easy for a
                                                compromised non-admin
                                                app to send a flood of
                                                huge UDP packets (with
                                                TCP you get throttled by
                                                the congestion
                                                controller).=C2=A0 So there
                                                WILL be DDoS protection
                                                devices out there to try
                                                to mitigate the problem,
                                                possibly by observing
                                                both directions of the
                                                flow and deciding
                                                whether a packet belongs
                                                to a valid flow or not.</p>
                                              <p class=3D"MsoNormal">=C2=A0=
</p>
                                              <p class=3D"MsoNormal">Since
                                                such middle boxes will
                                                be created, the more
                                                explicit and normative
                                                Invariants are about
                                                what one can expect, the
                                                less such middle boxes
                                                may decide for
                                                themselves.=C2=A0 For examp=
le
                                                (I did not think long
                                                about it), if some
                                                elements of path
                                                validation could land
                                                into Invariants
                                                (roughly, =E2=80=9Cno more =
than
                                                X packets/bytes can be
                                                sent on a new path w/o a
                                                return packet=E2=80=9D), a =
DDoS
                                                middle box may use this
                                                fact and active
                                                connection migration
                                                might still have a
                                                chance during an attack
                                                (NAT rebinding could be
                                                linked by DDoS boxes to
                                                an old connection via
                                                unchanged CID).</p>
                                              <p class=3D"MsoNormal">=C2=A0=
</p>
                                              <ul type=3D"disc">
                                                <li class=3D"MsoNormal">
                                                  Igor</li>
                                              </ul>
                                              <p class=3D"MsoNormal">=C2=A0=
</p>
                                              <p class=3D"MsoNormal">=C2=A0=
</p>
                                              <div style=3D"border-color:cu=
rrentcolor currentcolor currentcolor blue;border-style:none none none solid=
;border-width:medium medium medium 1.5pt;padding:0in 0in 0in 4pt">
                                                <p class=3D"MsoNormal" styl=
e=3D"margin-bottom:12pt"><b>From:</b>
                                                  Christian Huitema &lt;<a =
href=3D"mailto:huitema@huitema.net" target=3D"_blank">huitema@huitema.net</=
a>&gt;
                                                  <br>
                                                  <b>Sent:</b>
                                                  Wednesday, May 27,
                                                  2020 11:34 AM</p>
                                                <div>
                                                  <p class=3D"MsoNormal">On
                                                    5/27/2020 8:28 AM,
                                                    Kyle Rose wrote:</p>
                                                </div>
                                                <blockquote style=3D"margin=
-top:5pt;margin-bottom:5pt">
                                                  <p class=3D"MsoNormal">On
                                                    Wed, May 27, 2020 at
                                                    10:34 AM Ian Swett
                                                    &lt;ianswett=3D<a href=
=3D"mailto:40google.com@dmarc.ietf.org" target=3D"_blank">40google.com@dmar=
c.ietf.org</a>&gt;
                                                    wrote:</p>
                                                  <div>
                                                    <blockquote style=3D"bo=
rder-color:currentcolor currentcolor currentcolor rgb(204,204,204);border-s=
tyle:none none none solid;border-width:medium medium medium 1pt;padding:0in=
 0in 0in 6pt;margin:5pt 0in 5pt 4.8pt">
                                                      <div>
                                                        <p class=3D"MsoNorm=
al">I
                                                          was agreeing
                                                          with MT, but
                                                          I&#39;m happy
                                                          to=C2=A0see some
                                                          more MUSTs
                                                          added if
                                                          people feel
                                                          that&#39;d be
                                                          helpful.</p>
                                                      </div>
                                                    </blockquote>
                                                    <div>
                                                      <p class=3D"MsoNormal=
"><span style=3D"font-size:12pt">=C2=A0</span></p>
                                                      <div>
                                                        <div>
                                                          <p class=3D"MsoNo=
rmal"><span style=3D"font-size:12pt">Coincidentally, we were just talking a=
bout this
                                                          internally at
                                                          Akamai
                                                          yesterday.
                                                          IMO, an
                                                          invariants
                                                          document isn&#39;=
t
                                                          really helpful
                                                          if it isn&#39;t
                                                          normative, and
                                                          for it to be
                                                          normative it
                                                          (or a related
                                                          practices doc
                                                          for operators)
                                                          really needs
                                                          to spell out
                                                          clear
                                                          boundaries for
                                                          operators with
                                                          MUSTs..</span></p=
>
                                                        </div>
                                                        <div>
                                                          <p class=3D"MsoNo=
rmal"><span style=3D"font-size:12pt">=C2=A0</span></p>
                                                        </div>
                                                        <div>
                                                          <p class=3D"MsoNo=
rmal"><span style=3D"font-size:12pt">The example that came up yesterday was=
 around
                                                          operators
                                                          filtering QUIC
                                                          in the event
                                                          of a DDoS: one
                                                          recommendation
                                                          based on some
                                                          conversations
                                                          going back at
                                                          least to
                                                          Prague 2019
                                                          was to hash
                                                          packets on
                                                          4-tuple and
                                                          filter those
                                                          below a hash
                                                          value chosen
                                                          for a desired
                                                          ingress limit
                                                          instead of
                                                          doing what
                                                          most operators
                                                          do with UDP
                                                          today, which
                                                          is to cap UDP
                                                          throughput and
                                                          just drop
                                                          packets
                                                          randomly or
                                                          tail drop.</span>=
</p>
                                                        </div>
                                                      </div>
                                                    </div>
                                                  </div>
                                                </blockquote>
                                                <p>Interesting. Did they
                                                  consider using the
                                                  CID, or a fraction of
                                                  it? This looks
                                                  entirely like the
                                                  scenario for which we
                                                  developed stateless
                                                  retry.</p>
                                                <blockquote style=3D"margin=
-top:5pt;margin-bottom:5pt">
                                                  <div>
                                                    <div>
                                                      <div>
                                                        <div>
                                                          <p class=3D"MsoNo=
rmal"><span style=3D"font-size:12pt">=C2=A0</span></p>
                                                        </div>
                                                        <div>
                                                          <p class=3D"MsoNo=
rmal"><span style=3D"font-size:12pt">This recommendation certainly imposes =
some
                                                          constraints on
                                                          future
                                                          protocol
                                                          development
                                                          that motivate
                                                          new
                                                          invariants:
                                                          for instance,
                                                          it would
                                                          preclude
                                                          sharding a
                                                          connection
                                                          across
                                                          multiple
                                                          source ports
                                                          (not that
                                                          there is
                                                          necessarily a
                                                          reason to do
                                                          this; it&#39;s
                                                          just an
                                                          example). But
                                                          more
                                                          importantly,
                                                          it goes beyond
                                                          invariants:
                                                          it&#39;s one amon=
g
                                                          many practices
                                                          compatible
                                                          with the
                                                          current set of
                                                          invariants,
                                                          some
                                                          reasonable and
                                                          some terrible.</s=
pan></p>
                                                        </div>
                                                      </div>
                                                    </div>
                                                  </div>
                                                </blockquote>
                                                <p>This would break the
                                                  &quot;preferred address&q=
uot;
                                                  redirection. Preferred
                                                  address migration may
                                                  or may not be spelled
                                                  out in the invariants.</p=
>
                                                <blockquote style=3D"margin=
-top:5pt;margin-bottom:5pt">
                                                  <div>
                                                    <div>
                                                      <div>
                                                        <div>
                                                          <p class=3D"MsoNo=
rmal"><span style=3D"font-size:12pt">=C2=A0</span></p>
                                                        </div>
                                                        <div>
                                                          <p class=3D"MsoNo=
rmal"><span style=3D"font-size:12pt">Operators are going to do things to QU=
IC traffic,
                                                          so it would be
                                                          good to offer
                                                          them
                                                          recommendations
                                                          that are
                                                          compatible
                                                          with broad
                                                          deployability.</s=
pan></p>
                                                        </div>
                                                      </div>
                                                    </div>
                                                  </div>
                                                </blockquote>
                                                <p>=C2=A0</p>
                                                <p>Yes, we do need the
                                                  invariants for that.</p>
                                                <p>-- Christian Huitema</p>
                                              </div>
                                            </div>
                                          </div>
                                        </blockquote>
                                      </div>
                                    </div>
                                  </div>
                                </div>
                              </blockquote>
                            </div>
                          </div>
                        </div>
                      </blockquote>
                    </div>
                  </blockquote>
                </div>
              </div>
            </blockquote>
          </div>
        </blockquote>
      </div>
    </blockquote>
  </div>

</blockquote></div>

--00000000000000f2dc05a8797ac9--


From nobody Sat Jun 20 11:34:51 2020
Return-Path: <paul@redbarn.org>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 35C183A08BB for <quic@ietfa.amsl.com>; Sat, 20 Jun 2020 11:34:49 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, SPF_HELO_NONE=0.001, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id nNLMAvJnSIC5 for <quic@ietfa.amsl.com>; Sat, 20 Jun 2020 11:34:47 -0700 (PDT)
Received: from family.redbarn.org (family.redbarn.org [IPv6:2001:559:8000:cd::5]) (using TLSv1.2 with cipher ADH-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 75BC23A07EA for <quic@ietf.org>; Sat, 20 Jun 2020 11:34:46 -0700 (PDT)
Received: from linux-9daj.localnet (dhcp-166.access.rits.tisf.net [24.104.150.166]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (1024 bits) server-digest SHA256) (Client did not present a certificate) by family.redbarn.org (Postfix) with ESMTPSA id 25C9AB07D0 for <quic@ietf.org>; Sat, 20 Jun 2020 18:34:44 +0000 (UTC)
From: Paul Vixie <paul@redbarn.org>
To: IETF QUIC WG <quic@ietf.org>
Subject: Re: Is the invariants draft really standards track?
Date: Sat, 20 Jun 2020 18:34:41 +0000
Message-ID: <1621715.PBmxXT1aCC@linux-9daj>
Organization: none
In-Reply-To: <f9e2c611-bb4d-bc80-dfe3-e323a08bfc5b@huitema.net>
References: <CAM4esxQBqfrz24riPQA_VGKcGp_TzW0pqb97KfFMtNdW9pUfDg@mail.gmail.com> <CAM4esxRWVRjVhxyYuuzwDGq_wfTjQHkY6KHG2rEPErO2aHXA0w@mail.gmail.com> <f9e2c611-bb4d-bc80-dfe3-e323a08bfc5b@huitema.net>
MIME-Version: 1.0
Content-Transfer-Encoding: 7Bit
Content-Type: text/plain; charset="us-ascii"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/4bvh9jTOoaBA90TK5hP9YoCtWtU>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Sat, 20 Jun 2020 18:34:49 -0000

On Friday, 19 June 2020 23:51:39 UTC Christian Huitema wrote:
> ...
> 
> The DOS box does not have to worry about what kind of traffic is coming
> in. It just has to open a context for the 5 tuple, and check whether it
> sees 1-RTT packets coming back. And then maybe count the volume of 1-RTT
> packets coming back.
> 
> The worry is that one of the bots might start a legitimate connection,
> then disclose its five tuple to the rest of the botnet. The whole botnet
> can then spoof the 5 tuple that was just pin-holed by the DOS box. A
> simple "open-close" logic is thus not good enough. The DOS box must also
> enforce some kind of rate limiting per 5 tuple.
> 
> Which also means that if a botnet can predict the 5 tuple used by a
> legitimate connection and then spoof it, it can DOS it. Once you start
> digging that particular rabbit hole, the joy never stops...

pinholing based on outbound is the worst possible solution to DDoS, except for 
all the others. stateful firewalls of this kind create _almost_ as many 
problems as they solve, and would not be used if an alternative existed. this 
is a BCP38 problem not a QUIC problem. QUIC's only responsibility is to not 
make this known-bad situation worse.

there are two important ways that QUIC can avoid making the situation worse.

first, make connection mobility an outbound-first activity for initiators. the 
new 5-tuple would have to be negotiated in-band, and the first packet of the 
new flow would have to come from the initiator, to create the new pinhole.

second, make the new flow identifiable by the DOS box (or other filtering 
router such as a firewall box) using some low-entropy invariant. at the moment 
all we have to go on for detecting new outbound QUIC flows is a destination 
UDP port number, and that's both constraining and fragile.

-- 
Paul



From nobody Sun Jun 21 00:32:45 2020
Return-Path: <do_not_reply@mnot.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 604E53A0A0A for <quic@ietfa.amsl.com>; Sun, 21 Jun 2020 00:32:43 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, HTML_MESSAGE=0.001, RCVD_IN_MSPIKE_H3=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=mnot.net header.b=bcZT4gIl; dkim=pass (2048-bit key) header.d=messagingengine.com header.b=K8Msx6og
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 12ynYGxZwglh for <quic@ietfa.amsl.com>; Sun, 21 Jun 2020 00:32:41 -0700 (PDT)
Received: from out5-smtp.messagingengine.com (out5-smtp.messagingengine.com [66.111.4.29]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id ECC073A0A09 for <quic@ietf.org>; Sun, 21 Jun 2020 00:32:40 -0700 (PDT)
Received: from compute1.internal (compute1.nyi.internal [10.202.2.41]) by mailout.nyi.internal (Postfix) with ESMTP id 452835C00BA for <quic@ietf.org>; Sun, 21 Jun 2020 03:32:40 -0400 (EDT)
Received: from mailfrontend1 ([10.202.2.162]) by compute1.internal (MEProxy); Sun, 21 Jun 2020 03:32:40 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=mnot.net; h= content-type:mime-version:from:to:subject:message-id:date; s= fm3; bh=jucA21yt3x0LH3TUSx/eoCtSi8QDHZ9GvAB0XY9ns00=; b=bcZT4gIl 8K55Gdlm8Fw1fEjURusWmvRHOs2QsPUfS/K+gTp4mz5jmNsb4QnKyhEYTO07u9Dl DYmM/flT+D5985aA38Va2tCm7R7SDzHEk/mizhBdBYHMrll1YTLyNqSd43iAFPGN t1JbIW+cr0z1xChhPwbSxVr+8SDCE7PF1M9AEPX34fDVtMUGMuml6+bK7PiMMBDw /2rMApjPDb/5gOdS8qW5j7fZb8/sUtZqwnEPmMZ5K9MMUW2C7VvziMAtH1JPWk6+ pWpL2djBRx5drtikOLHV51H+7RrJs4jM87ljg8Qsrff43uLQyVNzpsBDv4iB0Inc 9zxYcH8oJOaI9A==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d= messagingengine.com; h=content-type:date:from:message-id :mime-version:subject:to:x-me-proxy:x-me-proxy:x-me-sender :x-me-sender:x-sasl-enc; s=fm3; bh=jucA21yt3x0LH3TUSx/eoCtSi8QDH Z9GvAB0XY9ns00=; b=K8Msx6og1ojllqpf3YEa40Ca6Bihj6tPacUhnbeYDGf7j 47r+vaNfE/AxEegUGtW9r951yu3H0tWNKZAcOoYT5+wLB8oAQeheNpDf8ZZpZ7Fh SSWnoep+yG0d0FAHEaSFbHBPcjqn+0Q7Uon8Ng8m41SkZQn/7oOzvyfpE0f+X7/P /AoBauD1K1zmUEcoZr8c/9vtNFu2AG5AXzw9iJGJyXdjaGKQQZ3ivr5znT34+sOp qoUFF8b608w5xHaRF5mku5DJ/mSwGU7JK7eZDRHEJlu4BgXrXXsGogCtVqgbxQWv WR7llw/mZOz/pBXJxc/8eqOzaXc2z3PXiAAikheXQ==
X-ME-Sender: <xms:GA3vXvORWYDlv3o6XDZyogqbC6kcq0apqhem1DPdrMQUha8KRhE4gw>
X-ME-Proxy-Cause: gggruggvucftvghtrhhoucdtuddrgeduhedrudejledguddvvdcutefuodetggdotefrod ftvfcurfhrohhfihhlvgemucfhrghsthforghilhdpqfgfvfdpuffrtefokffrpgfnqfgh necuuegrihhlohhuthemuceftddtnecunecujfgurheptggghffvufesrgdttdertddtje enucfhrhhomheptfgvphhoshhithhorhihucettghtihhvihhthicuufhumhhmrghrhicu uehothcuoeguohgpnhhothgprhgvphhlhiesmhhnohhtrdhnvghtqeenucggtffrrghtth gvrhhnpeekfedvudetjedvfeekheeiveeugfefhfetteevgeffkefffeetffdvleehudei teenucffohhmrghinhepghhithhhuhgsrdgtohhmnecukfhppeehvddrudegledrudehke drvddvheenucevlhhushhtvghrufhiiigvpedtnecurfgrrhgrmhepmhgrihhlfhhrohhm peguohgpnhhothgprhgvphhlhiesmhhnohhtrdhnvght
X-ME-Proxy: <xmx:GA3vXp-SPa1LPVwC9c50UNg0ji1JCtE-zOvAdrtUITePNTWpRVDKAA> <xmx:GA3vXuSf5hhUsamkX4DeOQJLroIv3H2KvjBR0-1DJSnJG_3KTlKAAQ> <xmx:GA3vXjtY7DAuJm_t8jRBqzKYnuY-OlC0KEmBix6o_qtT6u_o9IYbqQ> <xmx:GA3vXnlVSlOq9UMgrFa14YeeP8SecK-QXZWQ1h8EwB2_JR_pWrSCNA>
Received: from fv-az52.internal.cloudapp.net (unknown [52.149.158.225]) by mail.messagingengine.com (Postfix) with ESMTPA id DA8DC3280068 for <quic@ietf.org>; Sun, 21 Jun 2020 03:32:39 -0400 (EDT)
Content-Type: multipart/alternative; boundary="===============0517384627279222484=="
MIME-Version: 1.0
From: Repository Activity Summary Bot <do_not_reply@mnot.net>
To: quic@ietf.org
Subject: Weekly github digest (QUIC Activity Summary)
Message-Id: <20200621073239.DA8DC3280068@mailuser.nyi.internal>
Date: Sun, 21 Jun 2020 03:32:39 -0400 (EDT)
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/TLYSfNxydPGYguhmSIVaiP8XEm4>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Sun, 21 Jun 2020 07:32:43 -0000

--===============0517384627279222484==
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain; charset="utf-8"; format="flowed"




Events without label "editorial"

Issues
------
* quicwg/base-drafts (+7/-0/=F0=9F=92=AC25)
  7 issues created:
  - recovery: Clarifying the priority of Probe packet (by junhochoi)
    https://github.com/quicwg/base-drafts/issues/3780=20
  - Third ECN validation check and lost acknowledgments (by martinthomson)
    https://github.com/quicwg/base-drafts/issues/3778=20
  - Normative Text in Invariants (by martinduke)
    https://github.com/quicwg/base-drafts/issues/3773=20
  - DetectLostPackets still being referred to (by larseggert)
    https://github.com/quicwg/base-drafts/issues/3771=20
  - disable_active_migration with SPA (by igorlord)
    https://github.com/quicwg/base-drafts/issues/3765=20
  - [Editorial] Lots of forward references (by wbl)
    https://github.com/quicwg/base-drafts/issues/3763=20
  - [Editorial] Section 3.0 repeates with a must text from 2.1 (by wbl)
    https://github.com/quicwg/base-drafts/issues/3762=20

  5 issues received 25 new comments:
  - #3778 Third ECN validation check and lost acknowledgments (3 by janaiye=
ngar, martinthomson)
    https://github.com/quicwg/base-drafts/issues/3778 [-transport]=20
  - #3765 disable_active_migration with SPA (12 by MikeBishop, ianswett, ig=
orlord, kazuho)
    https://github.com/quicwg/base-drafts/issues/3765 [-transport]=20
  - #3763 [Editorial] Lots of forward references (1 by ianswett)
    https://github.com/quicwg/base-drafts/issues/3763 [-transport]=20
  - #3733 Question on ECN validation (8 by gloinul, larseggert, martinthoms=
on, mirjak)
    https://github.com/quicwg/base-drafts/issues/3733 [-transport]=20
  - #3722 QUIC Transport: Flow Control (1 by evolodina)
    https://github.com/quicwg/base-drafts/issues/3722 [-transport]=20

* quicwg/load-balancers (+2/-2/=F0=9F=92=AC0)
  2 issues created:
  - Add retry_source_connection_id to Retry Service Token Format (by martin=
duke)
    https://github.com/quicwg/load-balancers/issues/31=20
  - Routing of ICMP Packet too big messages (by martinduke)
    https://github.com/quicwg/load-balancers/issues/28=20

  2 issues closed:
  - Routing of ICMP Packet too big messages https://github.com/quicwg/load-=
balancers/issues/28=20
  - SCID acronym https://github.com/quicwg/load-balancers/issues/7=20



Pull requests
-------------
* quicwg/base-drafts (+5/-2/=F0=9F=92=AC2)
  5 pull requests submitted:
  - Try to make ECN validation more robust (by martinthomson)
    https://github.com/quicwg/base-drafts/pull/3779 [-transport] [design]=20
  - Make language in ECN validation clearer (by martinthomson)
    https://github.com/quicwg/base-drafts/pull/3777 [-transport] [design]=20
  - Added normative text to invariants (by martinduke)
    https://github.com/quicwg/base-drafts/pull/3774=20
  - ROBUST: ePrint publication (by fxguenther)
    https://github.com/quicwg/base-drafts/pull/3768=20
  - Some more explanation on ECN validation to address issue #3733 (by mirj=
ak)
    https://github.com/quicwg/base-drafts/pull/3767=20

  1 pull requests received 2 new comments:
  - #3779 Try to make ECN validation more robust (2 by janaiyengar, martint=
homson)
    https://github.com/quicwg/base-drafts/pull/3779 [-transport] [design]=20

  2 pull requests merged:
  - http: expand "ALPN" on first use and ref RFC 7301
    https://github.com/quicwg/base-drafts/pull/3758=20
  - ROBUST: ePrint publication
    https://github.com/quicwg/base-drafts/pull/3768=20

* quicwg/ops-drafts (+0/-0/=F0=9F=92=AC1)
  1 pull requests received 1 new comments:
  - #94 DDoS mitigations - linking after NAT rebinding and UDP Policing (1 =
by igorlord)
    https://github.com/quicwg/ops-drafts/pull/94=20

* quicwg/load-balancers (+3/-2/=F0=9F=92=AC0)
  3 pull requests submitted:
  - Add retry_source_connection_id to tokens (by martinduke)
    https://github.com/quicwg/load-balancers/pull/32=20
  - Update draft-ietf-quic-load-balancers.md (by martinduke)
    https://github.com/quicwg/load-balancers/pull/30=20
  - Update draft-ietf-quic-load-balancers.md (by martinduke)
    https://github.com/quicwg/load-balancers/pull/29=20

  2 pull requests merged:
  - ICMP Packet routing
    https://github.com/quicwg/load-balancers/pull/30=20
  - Get rid of SCID acronym
    https://github.com/quicwg/load-balancers/pull/26=20


Repositories tracked by this digest:
-----------------------------------
* https://github.com/quicwg/base-drafts
* https://github.com/quicwg/ops-drafts
* https://github.com/quicwg/datagram
* https://github.com/quicwg/load-balancers
* https://github.com/quicwg/version-negotiation

--===============0517384627279222484==
Content-Type: text/html; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable

<!doctype html>
<html lang=3D"en">
<head>
<meta charset=3D"utf-8">
<title>Weekly github digest (QUIC Activity Summary)</title>
<style>
body { font-family: Gotham, "Helvetica Neue", Helvetica, Arial, sans-serif;=
 font-size: 14px; }
h2 { margin-top: 3em; color: #A52A2A; font-style: italic; font-weight: norm=
al; }
h3 { margin-bottom:0; margin-top: 2em; font-size: 1.2em; }
h1+h2 { margin-top: 1em; }
a { color: #bb6219; text-decoration: none; }
li { margin-bottom: .35em; }
.repos { margin-bottom: 0; margin-top:0; line-height: 1.2; }
.new { color: red; }
.label { display: inline;
	padding: .2em .6em .3em;
	font-size: 75%;
	font-weight: 700;
	line-height: 1;
	color: #fff;
	text-align: center;
	white-space: nowrap;
	vertical-align: baseline;
	border-radius: .25em;
}
</style>
</head>

<body>
<h1>Sunday June 21, 2020</h1>

<p>Events without label "editorial"</p>

<h2>Issues</h2>

<h3>quicwg/base-drafts (+7/-0/=F0=9F=92=AC25)</h3>
  <p class=3D"new">7 issues created:</p>
  <ul>
  <li>#3780 <a href=3D"https://github.com/quicwg/base-drafts/issues/3780">r=
ecovery: Clarifying the priority of Probe packet</a> (by junhochoi) </li>
 =20
  <li>#3778 <a href=3D"https://github.com/quicwg/base-drafts/issues/3778">T=
hird ECN validation check and lost acknowledgments</a> (by martinthomson) <=
/li>
 =20
  <li>#3773 <a href=3D"https://github.com/quicwg/base-drafts/issues/3773">N=
ormative Text in Invariants</a> (by martinduke) </li>
 =20
  <li>#3771 <a href=3D"https://github.com/quicwg/base-drafts/issues/3771">D=
etectLostPackets still being referred to</a> (by larseggert) </li>
 =20
  <li>#3765 <a href=3D"https://github.com/quicwg/base-drafts/issues/3765">d=
isable_active_migration with SPA</a> (by igorlord) </li>
 =20
  <li>#3763 <a href=3D"https://github.com/quicwg/base-drafts/issues/3763">[=
Editorial] Lots of forward references</a> (by wbl) </li>
 =20
  <li>#3762 <a href=3D"https://github.com/quicwg/base-drafts/issues/3762">[=
Editorial] Section 3.0 repeates with a must text from 2.1</a> (by wbl) </li>
  </ul>

  <p>5 issues received 25 new comments:</p>
  <ul>
  <li>#3778 <a href=3D"https://github.com/quicwg/base-drafts/issues/3778">T=
hird ECN validation check and lost acknowledgments</a> (3 by janaiyengar, m=
artinthomson) <span class=3D"label" style=3D"background-color: #fbca04; col=
or: #000000">-transport</span> </li>
 =20
  <li>#3765 <a href=3D"https://github.com/quicwg/base-drafts/issues/3765">d=
isable_active_migration with SPA</a> (12 by MikeBishop, ianswett, igorlord,=
 kazuho) <span class=3D"label" style=3D"background-color: #fbca04; color: #=
000000">-transport</span> </li>
 =20
  <li>#3763 <a href=3D"https://github.com/quicwg/base-drafts/issues/3763">[=
Editorial] Lots of forward references</a> (1 by ianswett) <span class=3D"la=
bel" style=3D"background-color: #fbca04; color: #000000">-transport</span> =
</li>
 =20
  <li>#3733 <a href=3D"https://github.com/quicwg/base-drafts/issues/3733">Q=
uestion on ECN validation</a> (8 by gloinul, larseggert, martinthomson, mir=
jak) <span class=3D"label" style=3D"background-color: #fbca04; color: #0000=
00">-transport</span> </li>
 =20
  <li>#3722 <a href=3D"https://github.com/quicwg/base-drafts/issues/3722">Q=
UIC Transport: Flow Control</a> (1 by evolodina) <span class=3D"label" styl=
e=3D"background-color: #fbca04; color: #000000">-transport</span> </li>
  </ul>


<h3>quicwg/load-balancers (+2/-2/=F0=9F=92=AC0)</h3>
  <p class=3D"new">2 issues created:</p>
  <ul>
  <li>#31 <a href=3D"https://github.com/quicwg/load-balancers/issues/31">Ad=
d retry_source_connection_id to Retry Service Token Format</a> (by martindu=
ke) </li>
 =20
  <li>#28 <a href=3D"https://github.com/quicwg/load-balancers/issues/28">Ro=
uting of ICMP Packet too big messages</a> (by martinduke) </li>
  </ul>


  <p>2 issues closed:</p>
  <ul>
  <li>#28 <a href=3D"https://github.com/quicwg/load-balancers/issues/28">Ro=
uting of ICMP Packet too big messages</a> </li>
 =20
  <li>#7 <a href=3D"https://github.com/quicwg/load-balancers/issues/7">SCID=
 acronym</a> </li>
  </ul>



<h2>Pull requests</h2>
<h3>quicwg/base-drafts (+5/-2/=F0=9F=92=AC2)</h3>
  <p class=3D"new">5 pull requests submitted:</p>
  <ul>
  <li>#3779 <a href=3D"https://github.com/quicwg/base-drafts/pull/3779">Try=
 to make ECN validation more robust</a> (by martinthomson) <span class=3D"l=
abel" style=3D"background-color: #fbca04; color: #">-transport</span> <span=
 class=3D"label" style=3D"background-color: #1d76db; color: #">design</span=
> </li>
 =20
  <li>#3777 <a href=3D"https://github.com/quicwg/base-drafts/pull/3777">Mak=
e language in ECN validation clearer</a> (by martinthomson) <span class=3D"=
label" style=3D"background-color: #fbca04; color: #">-transport</span> <spa=
n class=3D"label" style=3D"background-color: #1d76db; color: #">design</spa=
n> </li>
 =20
  <li>#3774 <a href=3D"https://github.com/quicwg/base-drafts/pull/3774">Add=
ed normative text to invariants</a> (by martinduke) </li>
 =20
  <li>#3768 <a href=3D"https://github.com/quicwg/base-drafts/pull/3768">ROB=
UST: ePrint publication</a> (by fxguenther) </li>
 =20
  <li>#3767 <a href=3D"https://github.com/quicwg/base-drafts/pull/3767">Som=
e more explanation on ECN validation to address issue #3733</a> (by mirjak)=
 </li>
  </ul>

  <p>1 pull requests received 2 new comments:</p>
  <ul>
  <li>#3779 <a href=3D"https://github.com/quicwg/base-drafts/pull/3779">Try=
 to make ECN validation more robust</a> (2 by janaiyengar, martinthomson) <=
span class=3D"label" style=3D"background-color: #fbca04; color: #000000">-t=
ransport</span> <span class=3D"label" style=3D"background-color: #1d76db; c=
olor: #ffffff">design</span> </li>
  </ul>

  <p>2 pull requests merged:</p>
  <ul>
  <li>#3758 <a href=3D"https://github.com/quicwg/base-drafts/pull/3758">htt=
p: expand &quot;ALPN&quot; on first use and ref RFC 7301</a> </li>
 =20
  <li>#3768 <a href=3D"https://github.com/quicwg/base-drafts/pull/3768">ROB=
UST: ePrint publication</a> </li>
  </ul>

<h3>quicwg/ops-drafts (+0/-0/=F0=9F=92=AC1)</h3>

  <p>1 pull requests received 1 new comments:</p>
  <ul>
  <li>#94 <a href=3D"https://github.com/quicwg/ops-drafts/pull/94">DDoS mit=
igations - linking after NAT rebinding and UDP Policing</a> (1 by igorlord)=
 </li>
  </ul>


<h3>quicwg/load-balancers (+3/-2/=F0=9F=92=AC0)</h3>
  <p class=3D"new">3 pull requests submitted:</p>
  <ul>
  <li>#32 <a href=3D"https://github.com/quicwg/load-balancers/pull/32">Add =
retry_source_connection_id to tokens</a> (by martinduke) </li>
 =20
  <li>#30 <a href=3D"https://github.com/quicwg/load-balancers/pull/30">Upda=
te draft-ietf-quic-load-balancers.md</a> (by martinduke) </li>
 =20
  <li>#29 <a href=3D"https://github.com/quicwg/load-balancers/pull/29">Upda=
te draft-ietf-quic-load-balancers.md</a> (by martinduke) </li>
  </ul>


  <p>2 pull requests merged:</p>
  <ul>
  <li>#30 <a href=3D"https://github.com/quicwg/load-balancers/pull/30">ICMP=
 Packet routing</a> </li>
 =20
  <li>#26 <a href=3D"https://github.com/quicwg/load-balancers/pull/26">Get =
rid of SCID acronym</a> </li>
  </ul>


<h2>Repositories tracked by this digest:</h2>
<ul class=3D"repos">
  <li><a href=3D"https://github.com/quicwg/base-drafts">https://github.com/=
quicwg/base-drafts</a></li>
  <li><a href=3D"https://github.com/quicwg/ops-drafts">https://github.com/q=
uicwg/ops-drafts</a></li>
  <li><a href=3D"https://github.com/quicwg/datagram">https://github.com/qui=
cwg/datagram</a></li>
  <li><a href=3D"https://github.com/quicwg/load-balancers">https://github.c=
om/quicwg/load-balancers</a></li>
  <li><a href=3D"https://github.com/quicwg/version-negotiation">https://git=
hub.com/quicwg/version-negotiation</a></li>
  </ul>
</body>
</html>

--===============0517384627279222484==--


From nobody Mon Jun 22 08:33:50 2020
Return-Path: <mbishop@evequefou.be>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id BDDC23A0E0A for <quic@ietfa.amsl.com>; Mon, 22 Jun 2020 08:33:49 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.901
X-Spam-Level: 
X-Spam-Status: No, score=-1.901 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, RCVD_IN_MSPIKE_H2=-0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=evequefou.onmicrosoft.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id c9mpQHC6HckW for <quic@ietfa.amsl.com>; Mon, 22 Jun 2020 08:33:47 -0700 (PDT)
Received: from NAM10-DM6-obe.outbound.protection.outlook.com (mail-dm6nam10on2097.outbound.protection.outlook.com [40.107.93.97]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id BECF63A0E01 for <quic@ietf.org>; Mon, 22 Jun 2020 08:33:47 -0700 (PDT)
ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=Q9WL0P7APZWxE3P2m1m8aGR4Hb/CEiddrBwHQOXpj8/KQyzmqx2e6xYQN+KvkPlG/MV24WQ//JBeWkJpdecbvU1wbi/BdYePUemCeDgTUfQuMEmBZc0sA+pgujxK4wMP8TL709lRFgihjKjmUXyB2ueRyAffO4iz87G2qBjI/iQ/VWpRAkJ52IlDfMu35tqOefM+f2/S+lZs4I2++uwIAiqfpdnFMxsBaCfInz8YxLISjcIINA1l5c6ZBBvQZcf24LOr2tR7hfDlgYB1kUgSsCRP3PuaQZ8k7ti89HZS16rpUCSzT/RNTTAlS+BPmUCeTDRRtJ8Axb2WFxUmZMxxjQ==
ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com;  s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=t5p9zzJJzZy3yvVHhhWKkJTpdwTSdHhg8XIj/xVZ28E=; b=gSMsXfxiHzRhnLco5CBB40lu7r5CgDlaG8OklbYIvmoyydSP45EUM9MNyCk8jnF92Qy+VxB0lxCCpJ3VZUqITT8Ct6PWpw83hDJbKXnnj481qBqQChTjK1kUmxrVilErqluqT5bUESkUKMBeuHxVgQSFo7VCxTL+80aMMbeU2/eaXdixbCjlpnSSoWGRuQnGYA1Ezafn0pLV2T9RyiJv2i2XPsl38kXNL8M/e2odZMH8aY3xUaAP1nnFSAlZ/WavEpt6riZ5GhZH0yoa7YIoVHiOUOTgJnTzykScSIHFV6FTmTXLWzhXUBEFzf+pUaUL7RJxJHa2EgSErYa0s5Bvuw==
ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=evequefou.be; dmarc=pass action=none header.from=evequefou.be; dkim=pass header.d=evequefou.be; arc=none
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=evequefou.onmicrosoft.com; s=selector2-evequefou-onmicrosoft-com; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=t5p9zzJJzZy3yvVHhhWKkJTpdwTSdHhg8XIj/xVZ28E=; b=uIsJ11YdrZPfCmCJJlI8Xn7EXoJj+RhFhidu0AsBjK1cfO9QR+C+soEX8wZseroSxgRkxdwo54GEeyefZVT5+7X2r2jgsHWA+VeW8HOvzHaeRX7PriXCyTEfHFs86NXiO7tiWocEJe/4XWXI0lQ6YfgR0H3MTWE/am2TKXnAvGw=
Received: from CH2PR22MB2086.namprd22.prod.outlook.com (2603:10b6:610:8c::8) by CH2PR22MB2008.namprd22.prod.outlook.com (2603:10b6:610:89::15) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.3109.22; Mon, 22 Jun 2020 15:33:45 +0000
Received: from CH2PR22MB2086.namprd22.prod.outlook.com ([fe80::f0e6:8202:6056:a596]) by CH2PR22MB2086.namprd22.prod.outlook.com ([fe80::f0e6:8202:6056:a596%5]) with mapi id 15.20.3109.027; Mon, 22 Jun 2020 15:33:44 +0000
From: Mike Bishop <mbishop@evequefou.be>
To: Paul Vixie <paul@redbarn.org>, IETF QUIC WG <quic@ietf.org>
Subject: RE: Is the invariants draft really standards track?
Thread-Topic: Is the invariants draft really standards track?
Thread-Index: AQHWM4bluS9KwhzxrUO10xX7ZrRokKi7fymAgACBhgCAAA8vAIAAAZ4AgAAJJ4CAIYmPAIACi6CAgAA8RwCAACERgIAAEpsAgAAEloCAAAqGgIAABNkAgAAOXoCAATnFgIAC4xfg
Date: Mon, 22 Jun 2020 15:33:44 +0000
Message-ID: <CH2PR22MB20860DB081531EF5D3458EF9DA970@CH2PR22MB2086.namprd22.prod.outlook.com>
References: <CAM4esxQBqfrz24riPQA_VGKcGp_TzW0pqb97KfFMtNdW9pUfDg@mail.gmail.com> <CAM4esxRWVRjVhxyYuuzwDGq_wfTjQHkY6KHG2rEPErO2aHXA0w@mail.gmail.com> <f9e2c611-bb4d-bc80-dfe3-e323a08bfc5b@huitema.net> <1621715.PBmxXT1aCC@linux-9daj>
In-Reply-To: <1621715.PBmxXT1aCC@linux-9daj>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
authentication-results: redbarn.org; dkim=none (message not signed) header.d=none;redbarn.org; dmarc=none action=none header.from=evequefou.be;
x-originating-ip: [2600:2b00:9309:6a01:a435:b5c6:6ad3:1dd0]
x-ms-publictraffictype: Email
x-ms-office365-filtering-correlation-id: 9d4d16f3-2774-4632-df2f-08d816c1a65d
x-ms-traffictypediagnostic: CH2PR22MB2008:
x-microsoft-antispam-prvs: <CH2PR22MB20083F86216DD1B396CC6198DA970@CH2PR22MB2008.namprd22.prod.outlook.com>
x-ms-oob-tlc-oobclassifiers: OLM:8882;
x-forefront-prvs: 0442E569BC
x-ms-exchange-senderadcheck: 1
x-microsoft-antispam: BCL:0;
x-microsoft-antispam-message-info: RTbq++dgvzld9yOnjXpY7uzmdYaiUZkWA+mLFlW6Cz8SKL5A1ChHVn5hGc/rrTDoW81Xs0gPGvyZl6XPkOvVwYZruhRqwyrXhga9qfrzja0EqqMl8Li6yjnndwiBm+/qKhxKLNGYq7AAgrymSE/lQUOiv4mrFGgJRn237fTR6tTiViL8XcQZ2Zdbl/DCGKijf6GWYioZ3EVJVO6wgjApVXvDkk/AtDxg+f0/QUlGdgZVofWApbcSCFZ2Eb/+KLCBaY2ZlltNQTi/JldTxdMLUv33x/xNzI8ggzB77fYn87/0ZFFQ6mVwzHv9C8lbYQrkkidQiC8k0enpckL/qI/96w==
x-forefront-antispam-report: CIP:255.255.255.255; CTRY:; LANG:en; SCL:1; SRV:;  IPV:NLI; SFV:NSPM; H:CH2PR22MB2086.namprd22.prod.outlook.com; PTR:; CAT:NONE;  SFTY:; SFS:(396003)(39830400003)(136003)(376002)(346002)(366004)(508600001)(5660300002)(52536014)(66946007)(66446008)(64756008)(66476007)(66556008)(33656002)(76116006)(316002)(9686003)(55016002)(86362001)(2906002)(71200400001)(8676002)(110136005)(66574015)(53546011)(6506007)(83380400001)(186003)(8936002)(7696005); DIR:OUT; SFP:1102; 
x-ms-exchange-antispam-messagedata: 8/ZFY56pXk+Qxt635ZZKZdLgfUSsvXNLEo2tnWRP1TKTcT7SaGo10OpawPEd3/i+AdJqHZDq7c6G2bS8MuaEyQ8zJ1TRX10a0VyFTIIwxOpQT+MjoFQnl8XHKqdWIrbw1JZYiZJzRwXWCrEEe5mU4g4tQioiOIlBsCOFYIr1q+B5ZFovVmSigEXlN2ICjO6PfVCNivzfF4EH3Abf7yo4SSAORe1Dp9VrMMeOoytP7gF+VCGSaWhORnSEIL6kZZJfa9UYYCZBWRVUqS2OaJmix0ouVKrJay9VgnmgAhD6FIL/xOepo0+wju96JkxNjmA2+as3EfJ2V/Zv+6ak+y4q2IELVvo0MQ/2mzGZzUyQLpUBIQjLygQzNUn6OTDtEhzwskTxUesO2JYgqDRBzTRdXAz/JfTyT2wMMIcU3VXkhXiqX2urOt7HxspFryk6QecTcZxVAV/ow3nAG7BPZeoFyJQRm8rzFIP97+DB2r8gGnjEuW0Pwlx8MdsKE/XUhRrA8lIKiRa9HDI6j39esrj5tS2fxx88SoIdans4bJKl334=
x-ms-exchange-transport-forked: True
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
X-OriginatorOrg: evequefou.be
X-MS-Exchange-CrossTenant-Network-Message-Id: 9d4d16f3-2774-4632-df2f-08d816c1a65d
X-MS-Exchange-CrossTenant-originalarrivaltime: 22 Jun 2020 15:33:44.7080 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Hosted
X-MS-Exchange-CrossTenant-id: 41eaf50b-882d-47eb-8c4c-0b5b76a9da8f
X-MS-Exchange-CrossTenant-mailboxtype: HOSTED
X-MS-Exchange-CrossTenant-userprincipalname: 9Is5Xi8+ewW1RJNmISfxonc3CfoYSqsaw6IgSEgj2xd1rZxP1FrR4NlxbSBi00XdZobD1uVkG01nIkjwn5znFA==
X-MS-Exchange-Transport-CrossTenantHeadersStamped: CH2PR22MB2008
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/W0drgrftdyc1RDW6MINvEb00VXM>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 22 Jun 2020 15:33:50 -0000

Negotiating the 5-tuple in-band is problematic in the client-server directi=
on; in the presence of a NAT, the client doesn't know what address/port the=
 server is going to see when it uses a new path.  For the same reason (NAT =
opening), the client always goes first with connection migration, even if i=
t's the server's address changing.  However, we're assuming that the server=
 has the ability to provide the client a directly-reachable address, so tha=
t probably covers your in-band negotiation.

-----Original Message-----
From: QUIC <quic-bounces@ietf.org> On Behalf Of Paul Vixie
Sent: Saturday, June 20, 2020 2:35 PM
To: IETF QUIC WG <quic@ietf.org>
Subject: Re: Is the invariants draft really standards track?

On Friday, 19 June 2020 23:51:39 UTC Christian Huitema wrote:
> ...
>=20
> The DOS box does not have to worry about what kind of traffic is=20
> coming in. It just has to open a context for the 5 tuple, and check=20
> whether it sees 1-RTT packets coming back. And then maybe count the=20
> volume of 1-RTT packets coming back.
>=20
> The worry is that one of the bots might start a legitimate connection,=20
> then disclose its five tuple to the rest of the botnet. The whole=20
> botnet can then spoof the 5 tuple that was just pin-holed by the DOS=20
> box. A simple "open-close" logic is thus not good enough. The DOS box=20
> must also enforce some kind of rate limiting per 5 tuple.
>=20
> Which also means that if a botnet can predict the 5 tuple used by a=20
> legitimate connection and then spoof it, it can DOS it. Once you start=20
> digging that particular rabbit hole, the joy never stops...

pinholing based on outbound is the worst possible solution to DDoS, except =
for all the others. stateful firewalls of this kind create _almost_ as many=
 problems as they solve, and would not be used if an alternative existed. t=
his is a BCP38 problem not a QUIC problem. QUIC's only responsibility is to=
 not make this known-bad situation worse.

there are two important ways that QUIC can avoid making the situation worse=
.

first, make connection mobility an outbound-first activity for initiators. =
the new 5-tuple would have to be negotiated in-band, and the first packet o=
f the new flow would have to come from the initiator, to create the new pin=
hole.

second, make the new flow identifiable by the DOS box (or other filtering r=
outer such as a firewall box) using some low-entropy invariant. at the mome=
nt all we have to go on for detecting new outbound QUIC flows is a destinat=
ion UDP port number, and that's both constraining and fragile.

--
Paul



From nobody Mon Jun 22 10:26:36 2020
Return-Path: <paul@redbarn.org>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 343DD3A0FC4 for <quic@ietfa.amsl.com>; Mon, 22 Jun 2020 10:26:36 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, SPF_HELO_NONE=0.001, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id tLmt0aMBsICV for <quic@ietfa.amsl.com>; Mon, 22 Jun 2020 10:26:35 -0700 (PDT)
Received: from family.redbarn.org (family.redbarn.org [24.104.150.213]) (using TLSv1.2 with cipher ADH-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 545BE3A0FC2 for <quic@ietf.org>; Mon, 22 Jun 2020 10:26:35 -0700 (PDT)
Received: from linux-9daj.localnet (dhcp-166.access.rits.tisf.net [24.104.150.166]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (1024 bits) server-digest SHA256) (Client did not present a certificate) by family.redbarn.org (Postfix) with ESMTPSA id 35EE3B0588; Mon, 22 Jun 2020 17:26:32 +0000 (UTC)
From: Paul Vixie <paul@redbarn.org>
To: IETF QUIC WG <quic@ietf.org>, Mike Bishop <mbishop@evequefou.be>
Subject: Re: Is the invariants draft really standards track?
Date: Mon, 22 Jun 2020 17:26:31 +0000
Message-ID: <1865626.yaPRte9qB2@linux-9daj>
Organization: none
In-Reply-To: <CH2PR22MB20860DB081531EF5D3458EF9DA970@CH2PR22MB2086.namprd22.prod.outlook.com>
References: <CAM4esxQBqfrz24riPQA_VGKcGp_TzW0pqb97KfFMtNdW9pUfDg@mail.gmail.com> <1621715.PBmxXT1aCC@linux-9daj> <CH2PR22MB20860DB081531EF5D3458EF9DA970@CH2PR22MB2086.namprd22.prod.outlook.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 7Bit
Content-Type: text/plain; charset="us-ascii"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/FedLX6qn1qSbRptDgVhehk0ZZHg>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Mon, 22 Jun 2020 17:26:36 -0000

On Monday, 22 June 2020 15:33:44 UTC Mike Bishop wrote:
> Negotiating the 5-tuple in-band is problematic in the client-server
> direction; in the presence of a NAT, the client doesn't know what
> address/port the server is going to see when it uses a new path.  For the
> same reason (NAT opening), the client always goes first with connection
> migration, even if it's the server's address changing.  However, we're
> assuming that the server has the ability to provide the client a
> directly-reachable address, so that probably covers your in-band
> negotiation.

ty. 

-- 
Paul



From ryan_c_hamilton@apple.com  Mon Jun 22 18:42:22 2020
Return-Path: <ryan_c_hamilton@apple.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id DE2873A16CB for <quic@ietfa.amsl.com>; Mon, 22 Jun 2020 18:42:22 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.099
X-Spam-Level: 
X-Spam-Status: No, score=-2.099 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, HTML_MESSAGE=0.001, RCVD_IN_MSPIKE_H4=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=apple.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ZyHfGteo3iLU for <quic@ietfa.amsl.com>; Mon, 22 Jun 2020 18:42:21 -0700 (PDT)
Received: from ma1-aaemail-dr-lapp03.apple.com (ma1-aaemail-dr-lapp03.apple.com [17.171.2.72]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 77DDD3A16CA for <quic@ietf.org>; Mon, 22 Jun 2020 18:42:21 -0700 (PDT)
Received: from pps.filterd (ma1-aaemail-dr-lapp03.apple.com [127.0.0.1]) by ma1-aaemail-dr-lapp03.apple.com (8.16.0.42/8.16.0.42) with SMTP id 05N1bA7q038180 for <quic@ietf.org>; Mon, 22 Jun 2020 18:42:19 -0700
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=apple.com; h=from : content-type : mime-version : subject : message-id : date : to; s=20180706; bh=WL5e7XRFvr3BYfN/dKtXle1v2OOWI7TxeMZut7XO4Tg=; b=uUA05ivB/Li5gkmcUuXG35pGN+0UZ+aTrlyGdhdOQsR158KEtuGe0yi0lcbqyDy29UgM DDjcahg/uprMcK6Dr3nqGFWzV/zjKBgpJ3XPnJCGIOlNq/kMUwE/98ZRp10hIeiMEp+J 0gjwA2SS4fbwq+s/otxXnJOSoQlwFev7tcSp8zZ5OziKAfaxJU7XuvmxJbHImA56wEDK oSKqi6jXWmid5uUaLq7OEpZRLcmNy1uzDDvMjrJ+7NapzJntQJzIl1/lBLNrjBtDpKLu x5f8sFdksKMzlg7mRo6xKj9f5ZQs+khK40B3MCJDAJTFU+HYpJYIBMCxKgJEFFVMeP1a bA== 
Received: from rn-mailsvcp-mta-lapp03.rno.apple.com (rn-mailsvcp-mta-lapp03.rno.apple.com [10.225.203.151]) by ma1-aaemail-dr-lapp03.apple.com with ESMTP id 31shev17gk-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128 verify=NO) for <quic@ietf.org>; Mon, 22 Jun 2020 18:42:19 -0700
Received: from rn-mailsvcp-mmp-lapp01.rno.apple.com (rn-mailsvcp-mmp-lapp01.rno.apple.com [17.179.253.14]) by rn-mailsvcp-mta-lapp03.rno.apple.com (Oracle Communications Messaging Server 8.1.0.5.20200312 64bit (built Mar 12 2020)) with ESMTPS id <0QCC00LCEVEIWDA0@rn-mailsvcp-mta-lapp03.rno.apple.com> for quic@ietf.org; Mon, 22 Jun 2020 18:42:18 -0700 (PDT)
Received: from process_milters-daemon.rn-mailsvcp-mmp-lapp01.rno.apple.com by rn-mailsvcp-mmp-lapp01.rno.apple.com (Oracle Communications Messaging Server 8.1.0.5.20200312 64bit (built Mar 12 2020)) id <0QCC00200VE4AJ00@rn-mailsvcp-mmp-lapp01.rno.apple.com> for quic@ietf.org; Mon, 22 Jun 2020 18:42:18 -0700 (PDT)
X-Va-A: 
X-Va-T-CD: 5397336512ab7eda75e5eb07a83e0a2d
X-Va-E-CD: 818da2cd381b6a03ecaf53b7fa13d948
X-Va-R-CD: fdb6af67d2fe9fe233e4487530ac1bfe
X-Va-CD: 0
X-Va-ID: 63973fa0-31fa-4104-b458-0b99c6f8daaa
X-V-A: 
X-V-T-CD: 5397336512ab7eda75e5eb07a83e0a2d
X-V-E-CD: 818da2cd381b6a03ecaf53b7fa13d948
X-V-R-CD: fdb6af67d2fe9fe233e4487530ac1bfe
X-V-CD: 0
X-V-ID: b6922ba6-80ba-42a7-8c8c-b46680006741
X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.216, 18.0.687 definitions=2020-06-22_16:2020-06-22, 2020-06-22 signatures=0
Received: from [17.235.0.92] (unknown [17.235.0.92]) by rn-mailsvcp-mmp-lapp01.rno.apple.com (Oracle Communications Messaging Server 8.1.0.5.20200312 64bit (built Mar 12 2020)) with ESMTPSA id <0QCC011KTVE25300@rn-mailsvcp-mmp-lapp01.rno.apple.com> for quic@ietf.org; Mon, 22 Jun 2020 18:42:18 -0700 (PDT)
From: Ryan Hamilton <ryan_c_hamilton@apple.com>
Content-type: multipart/alternative; boundary="Apple-Mail=_B8C940C4-666C-4D0D-B48C-361362A954A9"
MIME-version: 1.0 (Mac OS X Mail 13.0 \(3623.0.8\))
Subject: Experimental HTTP/3 support in iOS 14 and macOS Big Sur
Message-id: <852238C5-BB41-4B98-B9BC-EB707831F23F@apple.com>
Date: Mon, 22 Jun 2020 18:42:18 -0700
To: quic@ietf.org
X-Mailer: Apple Mail (2.3623.0.8)
X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.216, 18.0.687 definitions=2020-06-22_16:2020-06-22, 2020-06-22 signatures=0
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/pDGunVWtGLi3CxBHfjkEDZhW4Cg>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 23 Jun 2020 02:10:57 -0000

--Apple-Mail=_B8C940C4-666C-4D0D-B48C-361362A954A9
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=utf-8

Howdy Folks,

We=E2=80=99re thrilled to announce that the beta releases of iOS 14 and =
macOS Big Sur include experimental support for HTTP/3! There are options =
to enable HTTP/3 in Safari as well as for development versions of your =
applications. On Friday, there is a WWDC session, Boost performance and =
security with modern networking =
<https://developer.apple.com/wwdc20/10111>, which includes more details. =
In the meantime, the iOS 14 beta Release Notes =
<https://developer.apple.com/documentation/ios-ipados-release-notes/ios-ip=
ados-14-beta-release-notes> and macOS Big Sur beta Release Notes =
<https://developer.apple.com/documentation/macos-release-notes/macos-big-s=
ur-11-beta-release-notes> contain instructions for enabling HTTP/3.

Currently, draft-27 is supported. Please try it out during these early =
betas and send us feedback before the general release. You can do this =
from Feedback Assistant <https://feedbackassistant.apple.com/> or in the =
#applequic Slack channel.

Cheers,

Ryan

--Apple-Mail=_B8C940C4-666C-4D0D-B48C-361362A954A9
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html;
	charset=utf-8

<html><head><meta http-equiv=3D"Content-Type" content=3D"text/html; =
charset=3Dutf-8"></head><body style=3D"word-wrap: break-word; =
-webkit-nbsp-mode: space; line-break: after-white-space;" class=3D""><span=
 style=3D"font-size: 14px;" class=3D""><font face=3D"Trebuchet MS" =
class=3D"">Howdy Folks,<br class=3D""><br class=3D"">We=E2=80=99re =
thrilled to announce that the beta releases of iOS 14 and macOS Big Sur =
include experimental support for HTTP/3! There are options to enable =
HTTP/3 in Safari as well as for development versions of your =
applications. On Friday, there is a WWDC session,&nbsp;<a =
href=3D"https://developer.apple.com/wwdc20/10111" class=3D"">Boost =
performance and security with modern networking</a>, which includes more =
details. In the meantime, the iOS 14 beta&nbsp;<a =
href=3D"https://developer.apple.com/documentation/ios-ipados-release-notes=
/ios-ipados-14-beta-release-notes" class=3D"">Release Notes</a>&nbsp;and =
macOS Big Sur beta&nbsp;<a =
href=3D"https://developer.apple.com/documentation/macos-release-notes/maco=
s-big-sur-11-beta-release-notes" class=3D"">Release =
Notes</a>&nbsp;contain instructions for enabling HTTP/3.<br class=3D""><br=
 class=3D"">Currently, draft-27 is supported. Please try it out during =
these early betas and send us feedback before the general release. You =
can do this from&nbsp;<a href=3D"https://feedbackassistant.apple.com/" =
class=3D"">Feedback Assistant</a>&nbsp;or in the #applequic Slack =
channel.<br class=3D""><br class=3D"">Cheers,<br class=3D""><br =
class=3D"">Ryan</font></span><br class=3D""></body></html>=

--Apple-Mail=_B8C940C4-666C-4D0D-B48C-361362A954A9--


From nobody Tue Jun 23 16:15:55 2020
Return-Path: <jri.ietf@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id D5F973A0C2E for <quic@ietfa.amsl.com>; Tue, 23 Jun 2020 16:15:53 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.837
X-Spam-Level: 
X-Spam-Status: No, score=-1.837 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, HTML_OBFUSCATE_05_10=0.26, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id xTnatZIcl7ca for <quic@ietfa.amsl.com>; Tue, 23 Jun 2020 16:15:52 -0700 (PDT)
Received: from mail-lf1-x132.google.com (mail-lf1-x132.google.com [IPv6:2a00:1450:4864:20::132]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 1D1603A0C2D for <quic@ietf.org>; Tue, 23 Jun 2020 16:15:52 -0700 (PDT)
Received: by mail-lf1-x132.google.com with SMTP id g139so238756lfd.10 for <quic@ietf.org>; Tue, 23 Jun 2020 16:15:52 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=gHdojI/DRLmiYvl1U7ELyJU/dVps2rfysgjrugObF50=; b=NZblqXKXg7N7499VwuMhG8w71B5cupfptuWQRRKMY9eSswv8ulY3ZR1yLV+NXYWxBj GLZRiUsNYqvTA8ldBZWDK4V/J+uJhpI66RHuQ6bBNpCoTPra3vz5xj99rYk13NkDesRK hbUUgI0ty8TZ+uH2GGXLm5z0FgaqQFy/Mt03Zcu9mr4F7zqJq4Ug/pyBsa2KRpQGO31z HCuYuM4Etnb1Nnx0J/zNnTONGhv/pyt2hwM6AgXFccofW16YGGCRsi5GeBA2K/+2Pqar ycyCItkyv+wpFsxDx35FRNj4G5qhI7k1L+ySKxrgmwKpGfoybNHirScITx99dwlQHDYR E3/g==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=gHdojI/DRLmiYvl1U7ELyJU/dVps2rfysgjrugObF50=; b=jU0PYJ2jUJ2IBNtbRJSjaKuR16SDWvA9ksobjNSF1bSUzg6Odk0ljRaT+mFts21IAs 3sxPm9/Nz4KgoYVHZ6D40nCUx2zW3ymKQhzb7bAj27XhVYdbY9cOWc9B9DAJ8/Xt0HAL rM7zqHVv0NT4plmkJCjd1LVAgB6+0m951A18qiDgFWG711oSvx0kWx5/z224lILERg6n SpPNFFVz+seZt51QWsbbb8HriDLg8zq+XyDsdI/i1djNmzM2nPMk3lJxV6dn0FWQHRtn 4K07QRHom/HcmZZ43Hupbpq5VUPnJWbC3BB7Sn/Bpo+K0zoXJ5KckB6EgIy2tQr/qeNw EhMw==
X-Gm-Message-State: AOAM531O9orsgfV9/QFmEPUQBuYs6rQ2DRGMDG8wmhjcKfX5OTJIL5pd +ql0cnAmJ8AMMaamDoWWIia/ZmW8LplYhB7aFms=
X-Google-Smtp-Source: ABdhPJw4e9iuejqi9bbJBzfaV8DeUHIb2AhO9MowsNm42k2YPM7xU9PyFVbJ4/M1tDqDIalu1H95GPRFGefo/db3seM=
X-Received: by 2002:a19:2209:: with SMTP id i9mr13743799lfi.46.1592954150209;  Tue, 23 Jun 2020 16:15:50 -0700 (PDT)
MIME-Version: 1.0
References: <852238C5-BB41-4B98-B9BC-EB707831F23F@apple.com>
In-Reply-To: <852238C5-BB41-4B98-B9BC-EB707831F23F@apple.com>
From: Jana Iyengar <jri.ietf@gmail.com>
Date: Tue, 23 Jun 2020 16:15:39 -0700
Message-ID: <CACpbDcf6wfKrA8Q0-CM5g7d6hx9MO7rBtbYUMfLmK-2EkHJ2gQ@mail.gmail.com>
Subject: Re: Experimental HTTP/3 support in iOS 14 and macOS Big Sur
To: Ryan Hamilton <ryan_c_hamilton=40apple.com@dmarc.ietf.org>
Cc: QUIC WG <quic@ietf.org>
Content-Type: multipart/alternative; boundary="000000000000444aa605a8c88a80"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/7KDd-Rmjs2klyVBfVTjtS7JhX_Y>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 23 Jun 2020 23:15:54 -0000

--000000000000444aa605a8c88a80
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Ryan,

Thank you for the excellent news! Do you have any sense of when draft-28/29
might be supported?

- jana

On Mon, Jun 22, 2020 at 7:11 PM Ryan Hamilton <ryan_c_hamilton=3D
40apple.com@dmarc.ietf.org> wrote:

> Howdy Folks,
>
> We=E2=80=99re thrilled to announce that the beta releases of iOS 14 and m=
acOS Big
> Sur include experimental support for HTTP/3! There are options to enable
> HTTP/3 in Safari as well as for development versions of your applications=
.
> On Friday, there is a WWDC session, Boost performance and security with
> modern networking <https://developer.apple.com/wwdc20/10111>, which
> includes more details. In the meantime, the iOS 14 beta Release Notes
> <https://developer.apple.com/documentation/ios-ipados-release-notes/ios-i=
pados-14-beta-release-notes> and
> macOS Big Sur beta Release Notes
> <https://developer.apple.com/documentation/macos-release-notes/macos-big-=
sur-11-beta-release-notes> contain
> instructions for enabling HTTP/3.
>
> Currently, draft-27 is supported. Please try it out during these early
> betas and send us feedback before the general release. You can do this fr=
om Feedback
> Assistant <https://feedbackassistant.apple.com/> or in the #applequic
> Slack channel.
>
> Cheers,
>
> Ryan
>

--000000000000444aa605a8c88a80
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div>Ryan,</div><div><br></div>Thank you for the excellent=
 news! Do you have any sense of when draft-28/29 might be supported?<div><b=
r></div><div>- jana</div></div><br><div class=3D"gmail_quote"><div dir=3D"l=
tr" class=3D"gmail_attr">On Mon, Jun 22, 2020 at 7:11 PM Ryan Hamilton &lt;=
ryan_c_hamilton=3D<a href=3D"mailto:40apple.com@dmarc.ietf.org">40apple.com=
@dmarc.ietf.org</a>&gt; wrote:<br></div><blockquote class=3D"gmail_quote" s=
tyle=3D"margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);pad=
ding-left:1ex"><div style=3D"overflow-wrap: break-word;"><span style=3D"fon=
t-size:14px"><font face=3D"Trebuchet MS">Howdy Folks,<br><br>We=E2=80=99re =
thrilled to announce that the beta releases of iOS 14 and macOS Big Sur inc=
lude experimental support for HTTP/3! There are options to enable HTTP/3 in=
 Safari as well as for development versions of your applications. On Friday=
, there is a WWDC session,=C2=A0<a href=3D"https://developer.apple.com/wwdc=
20/10111" target=3D"_blank">Boost performance and security with modern netw=
orking</a>, which includes more details. In the meantime, the iOS 14 beta=
=C2=A0<a href=3D"https://developer.apple.com/documentation/ios-ipados-relea=
se-notes/ios-ipados-14-beta-release-notes" target=3D"_blank">Release Notes<=
/a>=C2=A0and macOS Big Sur beta=C2=A0<a href=3D"https://developer.apple.com=
/documentation/macos-release-notes/macos-big-sur-11-beta-release-notes" tar=
get=3D"_blank">Release Notes</a>=C2=A0contain instructions for enabling HTT=
P/3.<br><br>Currently, draft-27 is supported. Please try it out during thes=
e early betas and send us feedback before the general release. You can do t=
his from=C2=A0<a href=3D"https://feedbackassistant.apple.com/" target=3D"_b=
lank">Feedback Assistant</a>=C2=A0or in the #applequic Slack channel.<br><b=
r>Cheers,<br><br>Ryan</font></span><br></div></blockquote></div>

--000000000000444aa605a8c88a80--


From nobody Wed Jun 24 16:48:16 2020
Return-Path: <caw@heapingbits.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 5A5EA3A11DE for <quic@ietfa.amsl.com>; Wed, 24 Jun 2020 16:48:14 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.098
X-Spam-Level: 
X-Spam-Status: No, score=-2.098 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_MSPIKE_H4=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=heapingbits.net header.b=c+HTKuk2; dkim=pass (2048-bit key) header.d=messagingengine.com header.b=chxLyS5K
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ov6mbdhPexJj for <quic@ietfa.amsl.com>; Wed, 24 Jun 2020 16:48:12 -0700 (PDT)
Received: from wout4-smtp.messagingengine.com (wout4-smtp.messagingengine.com [64.147.123.20]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 466EE3A11DD for <quic@ietf.org>; Wed, 24 Jun 2020 16:48:12 -0700 (PDT)
Received: from compute1.internal (compute1.nyi.internal [10.202.2.41]) by mailout.west.internal (Postfix) with ESMTP id 2768CA13 for <quic@ietf.org>; Wed, 24 Jun 2020 19:48:10 -0400 (EDT)
Received: from imap4 ([10.202.2.54]) by compute1.internal (MEProxy); Wed, 24 Jun 2020 19:48:10 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=heapingbits.net; h=mime-version:message-id:date:from:to:subject:content-type :content-transfer-encoding; s=fm1; bh=bfhPB6jUEB7yvAPtRGTYbzWGim KmRQ8KFdbHhv+XqDM=; b=c+HTKuk2KKKny0AitmeJVj0VjSxzzIa4/jpvuOPa6m dk61B/kaUCepcJzpyb1T0dDhgy1x/Cbpai3a3H0SS3bb/oekQdIgSjR2gyyCRBVT yz3eoE7ikKp2zFIYQNi0wY8BSZuycntK3Lw3NbRbaoQVijI3GM+o6TLdOCZpLrYg i4eMQ4m16zTowljHShKi6bzzV9kxo2iPe/sZeLzJ+/XJBYDqWuTKhM4pgJvsqSEf +v/mRyMVt6PEB0WJx45u3u85C1oGDnh/N4UTNvAOg6W8pXsSflyVLiZGUO/hqgMn HaKg+p3q1RmsIB1YfK29n4NMJu2lxRuZeMF7PZgf9AcA==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d= messagingengine.com; h=content-transfer-encoding:content-type :date:from:message-id:mime-version:subject:to:x-me-proxy :x-me-proxy:x-me-sender:x-me-sender:x-sasl-enc; s=fm3; bh=bfhPB6 jUEB7yvAPtRGTYbzWGimKmRQ8KFdbHhv+XqDM=; b=chxLyS5KMb1zp1S5wtM703 jxfaVTtF0+LrpuBdo/BntVedx3xksITEccRY/nwZY+e5r956iif1SyHRB7/LtngP 6AEgprg6gxJ53a8W2zFyPKHzpIAbApGTMmW/geCJBR/UGt20Ypkz+zPGReSRHrgL Hc/Utr5Y/R/T5tQimITqOsFzcQkQHCsHar+7MgmoUzy5eYVG1OVHs4CotCm1te0F YJczQxOofU50mibpXXZa4SoilN6AAi6bTOzDmz9c7TRhxunnnrpxmyVpfw+b6/cB NAHL5rf+PKScf59/9hzmdtOnEEaYQA4U+j2rIb2OcKnKwSKBqFy8eBnENkAOt4NQ ==
X-ME-Sender: <xms:OebzXvuKSqbn6AnkhejXBsBP5jRBCCaZG82OYVZYTMiW65zZOFRAAg>
X-ME-Proxy-Cause: gggruggvucftvghtrhhoucdtuddrgeduhedrudekkedgvdeiucetufdoteggodetrfdotf fvucfrrhhofhhilhgvmecuhfgrshhtofgrihhlpdfqfgfvpdfurfetoffkrfgpnffqhgen uceurghilhhouhhtmecufedttdenucenucfjughrpefofgggkfffhffvufgtgfesthhqre dtreerjeenucfhrhhomhepfdevhhhrihhsthhophhhvghrucghohhougdfuceotggrfies hhgvrghpihhnghgsihhtshdrnhgvtheqnecuggftrfgrthhtvghrnhepkeehvdehffegje dukedvfeegudevkefftdektdegveffvdeuuefgtdekveduledunecuffhomhgrihhnpehg ihhthhhusgdrtghomhdpihgrtghrrdhorhhgpdhsphhrihhnghgvrhdrtghomhdprhhhuh hlrdgrtgdruhhkpdhutghsugdrvgguuhdprggtmhdrohhrghenucevlhhushhtvghrufhi iigvpedtnecurfgrrhgrmhepmhgrihhlfhhrohhmpegtrgifsehhvggrphhinhhgsghith hsrdhnvght
X-ME-Proxy: <xmx:OebzXgfL9NfpvqZLKjK_gGLDOmamh2JQ9Fb4LaSEClGd2c7fT3NQZQ> <xmx:OebzXiwUdxWZWwXwujRSKC-8Qwfffo46R-1KTnQSzT5z_kOQpy8gRw> <xmx:OebzXuOKRiXCdvWuqab9qbi6xA4374kk2id3gCIaH7L2CUtaH2bpzg> <xmx:OebzXtcgahYi96hOKK03nXbdDcew_TlO53NUAyHIaPRkNH2I-VmjWQ>
Received: by mailuser.nyi.internal (Postfix, from userid 501) id 5931B3C00A1; Wed, 24 Jun 2020 19:48:09 -0400 (EDT)
X-Mailer: MessagingEngine.com Webmail Interface
User-Agent: Cyrus-JMAP/3.3.0-dev0-543-gda70334-fm-20200618.004-gda703345
Mime-Version: 1.0
Message-Id: <43bb1525-0afa-4c4e-a234-f6ccfacbdd29@www.fastmail.com>
Date: Wed, 24 Jun 2020 16:47:40 -0700
From: "Christopher Wood" <caw@heapingbits.net>
To: quic@ietf.org
Subject: New confidentiality and integrity limits
Content-Type: text/plain;charset=utf-8
Content-Transfer-Encoding: quoted-printable
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/6ABngnMhcwvIQQQ-zYC_Ok-0NaM>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 24 Jun 2020 23:48:14 -0000

Hi folks,=20

I'm following up here based on an issue and proposed PR I just filed aga=
inst against draft-ietf-quic-tls.

   Issue #3788: New confidentiality and integrity limits [https://github=
.com/quicwg/base-drafts/issues/3788
   PR: https://github.com/quicwg/base-drafts/pull/3789

TL;DR: We propose changing per-key limits to per-connection limits, with=
 different margins.=20

Ongoing analysis efforts with Jean Paul Degabriele, Felix G=C3=BCnther, =
Kenny Paterson, and Martin Thomson revealed that QUIC should consider ch=
anging its confidentiality and integrity limits to account for multi-use=
r security settings. As some background, the existing limits in draft-ie=
tf-quic-tls apply to a single key. They are derived from several existin=
g analyses in the literature, including [1], [2], [3], and [4]. Concrete=
ly, they seek to bound an attacker=E2=80=99s advantage in breaking the s=
ecurity guarantees of the AEAD algorithm for a specific key in question.=
 When a Key Update occurs, endpoints receive a new key, which resets the=
 encryption and forgery attempt counters for this new key.=20

This analysis fails to account for settings in which there are multiple,=
 independent users of an AEAD algorithm, each with their own unique key.=
 In these so-called =E2=80=9Cmulti-user=E2=80=9D settings, the attacker =
is allowed to perform some amount of offline work to help accelerate any=
 attack on the AEAD algorithm. However, rather than target a single user=
=E2=80=99s specific key, they are tasked with distinguishing all traffic=
 (under all keys) from an equal amount of random bits. In the public-key=
 setting, Bellare et al. [5] prove that the success probability in attac=
king one of the many independent users (or keys) is bounded by the succe=
ss probability of attacking a single user (or key) multiplied by the num=
ber of users present. (This result extends to the symmetric-key setting.=
)

Why is this important? Each key used throughout the duration of a single=
 connection is effectively a distinct user. This means that the attacker=
=E2=80=99s probability of success against any key in a connection must b=
e considered across all keys. Or, more precisely, the integrity limit co=
unters MUST NOT reset across Key Updates. (Confidentiality limits still =
apply for individual keys, rather than across Key Updates, since there=E2=
=80=99s no functional difference in the analyses between tearing down a =
connection and performing an update. The analyses consider encryption as=
 a function of plaintext blocks protected. Thus, users that need to send=
 N blocks of data will send N blocks of data, whether it be in one conne=
ction with many keys, or many connections with a single key.)

Fortunately, this may not be much of a problem in practice. Hoang et al.=
 [6] provide tight multi-user security bounds for AES-GCM with nonce ran=
domization (as is used by TLS 1.3 and QUIC), and those bounds permit a h=
igher amount than previously established for a single key. However, to t=
he best of our knowledge, there are no multi-user security analyses givi=
ng tighter bounds than the generic analyses of AEAD_CHACHA20_POLY1305 an=
d AEAD_AES_128_CCM. Thus, for the time being, we are stuck with the sing=
le-user integrity limits spread across Key Updates for these AEADs. Reco=
gnizing that this is a relatively new area for research, we are aiming t=
o be quite conservative in setting limits, though we do allow an attacke=
r a greater advantage in this multi-user attack than we would for a sing=
le connection. We may relax these limits if and when future analyses dem=
onstrate that it=E2=80=99s safe to do so.

Note that this is not an ideal outcome. A =E2=80=9Ctrue=E2=80=9D multi-u=
ser setting would consider all users, i.e., all keys in all connections =
used by all QUIC clients and servers, simultaneously. However, endpoints=
 cannot obtain such a global view of the Internet, and thus cannot make =
realistic parameter choices for bounds based on the number of users. The=
 compromise struck in this PR is to consider a narrow view of multi-user=
 security, i.e., one in which the only =E2=80=9Cusers=E2=80=9D in a conn=
ection are those introduced by Key Update messages. (This is certainly a=
 gap in the literature worth exploring in the future.) We also recommend=
 setting very strong targets for attacker advantage for a single key, wh=
ich we estimate will ensure that an attacker still has limited advantage=
 in the multi-user setting.

Felix, Martin, and I did our best to document the basis for the analysis=
 in this issue (and corresponding PR) in this CFRG I-D:

   https://github.com/chris-wood/draft-wood-cfrg-aead-limits

We believe the foundations for this analysis are sound, though we are of=
 course happy to learn if we made arithmetic mistakes. :-)

Thanks,
Chris, on behalf of Jean Paul, Felix, Kenny, and Martin

[1] https://eprint.iacr.org/2012/438.pdf
[2] https://eprint.iacr.org/2014/613.pdf
[3] https://link.springer.com/chapter/10.1007/3-540-36492-7_7
[4] https://www.isg.rhul.ac.uk/~kp/TLS-AEbounds.pdf
[5] https://cseweb.ucsd.edu/~mihir/papers/musu.pdf
[6] https://dl.acm.org/doi/10.1145/3243734.3243816


From nobody Wed Jun 24 17:58:37 2020
Return-Path: <mt@lowentropy.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 83F253A120B for <quic@ietfa.amsl.com>; Wed, 24 Jun 2020 17:58:34 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.098
X-Spam-Level: 
X-Spam-Status: No, score=-2.098 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_MSPIKE_H4=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=lowentropy.net header.b=OTkuYu3L; dkim=pass (2048-bit key) header.d=messagingengine.com header.b=T/hg1eGx
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ht7zzuG3djJN for <quic@ietfa.amsl.com>; Wed, 24 Jun 2020 17:58:32 -0700 (PDT)
Received: from wout4-smtp.messagingengine.com (wout4-smtp.messagingengine.com [64.147.123.20]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 8A7463A120A for <quic@ietf.org>; Wed, 24 Jun 2020 17:58:32 -0700 (PDT)
Received: from compute2.internal (compute2.nyi.internal [10.202.2.42]) by mailout.west.internal (Postfix) with ESMTP id 4AC0FA5F for <quic@ietf.org>; Wed, 24 Jun 2020 20:58:31 -0400 (EDT)
Received: from imap2 ([10.202.2.52]) by compute2.internal (MEProxy); Wed, 24 Jun 2020 20:58:31 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=lowentropy.net; h=mime-version:message-id:in-reply-to:references:date:from:to :subject:content-type:content-transfer-encoding; s=fm2; bh=dMwkg IFCmndv1dhVMP7p+qL68Z9W/x7q6O16Icg36MI=; b=OTkuYu3LM2le6xdw8XxH0 3SWV/HQBtjA6tltOTaW559j5go5xaT+9m+T2c2aZdqra/JLxg9BEUFNGRUQR8wPH r1sZevVLd4+zUrDSJGpYir74MZIWfBsE0HINVzyZqLgMkIxTeB1t3FE09LFCRszJ UIBJMqNqPvuT0SmdreL6+NML0b3qOrMD29dTvj2sGe5e29RSJ6AoxDqu8eHfK8jE anSb82HTLcdv5RlFQDiW/OzRsXz50YN0vh358j/ad8YrVUaFKewjrsPHWE/uPkVE 96bd4t3SsmmzZVfPeSEhKohRNIxi3lfUePPZj9tJYU5ILeEQsCm3Ljcp/cdUlqeZ A==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d= messagingengine.com; h=content-transfer-encoding:content-type :date:from:in-reply-to:message-id:mime-version:references :subject:to:x-me-proxy:x-me-proxy:x-me-sender:x-me-sender :x-sasl-enc; s=fm3; bh=dMwkgIFCmndv1dhVMP7p+qL68Z9W/x7q6O16Icg36 MI=; b=T/hg1eGx7mVliyAm8CKLPxwVzlNGWuEPsCv/sAiVfMW+6QSEdG6LMiErt ybFdl+Z2Uz3JmxmZU3SlRc3V9XVWBuxwt72Ol+6AswDKQ8gqDHNKQgO3Jt9Y7pri XUMK/JVETNDmmwh6JgzRk1xn/OyOk/GsAKnhtlMnfr7VMmS3vITUXQjo1yTVm7DF 2/1S/fc6+eUyYOJLy1cWQbWoKf36IZDzKCAcBls5Px8wbfaO0Z3hqdzya/RmJMwc RPcpmzvjVqAsLUzPlPJhdf3Ghhxc49WaPzSpUgU3+0kpegjEdAy7owDZWoTRTSG4 OvYExsiAy+VJhr4sG6q6jF2bG1fnw==
X-ME-Sender: <xms:tvbzXlacXT1WHZldTqdIk_UFEKfX-CdiIlhHPMQ6xmVqxwym6QvbuQ>
X-ME-Proxy-Cause: gggruggvucftvghtrhhoucdtuddrgeduhedrudekkedggedtucetufdoteggodetrfdotf fvucfrrhhofhhilhgvmecuhfgrshhtofgrihhlpdfqfgfvpdfurfetoffkrfgpnffqhgen uceurghilhhouhhtmecufedttdenucenucfjughrpefofgggkfgjfhffhffvufgtgfesth hqredtreerjeenucfhrhhomhepfdforghrthhinhcuvfhhohhmshhonhdfuceomhhtsehl ohifvghnthhrohhphidrnhgvtheqnecuggftrfgrthhtvghrnhepkeejgfetvdeuheevvd elgffgleegfeehueeufefhgeevgfeuffeggedvtdevkefgnecuffhomhgrihhnpehgihht hhhusgdrrdgtohhmpdhgihhthhhusgdrtghomhdpihgrtghrrdhorhhgpdhsphhrihhngh gvrhdrtghomhdprhhhuhhlrdgrtgdruhhkpdhutghsugdrvgguuhdprggtmhdrohhrghen ucevlhhushhtvghrufhiiigvpedtnecurfgrrhgrmhepmhgrihhlfhhrohhmpehmtheslh hofigvnhhtrhhophihrdhnvght
X-ME-Proxy: <xmx:tvbzXsbwjjZX387lHUa41-pgtOFwnCGvjXx0QwEkj2ljaOQsySogcw> <xmx:tvbzXn_AFkkj6FMDJaLMkikOb2fvYZ_N4Pxg3622Tdo_8zHsJLlgzA> <xmx:tvbzXjp8mWvM5_bUyuQbfy13g-n9luqHbY9e37IgdFn0Xl8A_Qqf7Q> <xmx:tvbzXp5wb3X8C3DmnCuLFA_wdB7uHD-0mkVRqruo3CwjXo-lFuZldQ>
Received: by mailuser.nyi.internal (Postfix, from userid 501) id A48CBE00A8; Wed, 24 Jun 2020 20:58:30 -0400 (EDT)
X-Mailer: MessagingEngine.com Webmail Interface
User-Agent: Cyrus-JMAP/3.3.0-dev0-543-gda70334-fm-20200618.004-gda703345
Mime-Version: 1.0
Message-Id: <e4cd18e8-0ffb-44c2-928b-d3cbbf50e8ec@www.fastmail.com>
In-Reply-To: <43bb1525-0afa-4c4e-a234-f6ccfacbdd29@www.fastmail.com>
References: <43bb1525-0afa-4c4e-a234-f6ccfacbdd29@www.fastmail.com>
Date: Thu, 25 Jun 2020 10:58:11 +1000
From: "Martin Thomson" <mt@lowentropy.net>
To: quic@ietf.org
Subject: Re: New confidentiality and integrity limits
Content-Type: text/plain;charset=utf-8
Content-Transfer-Encoding: quoted-printable
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/JYDW5Ey2oLX48Bj911Yh59NjOhk>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 25 Jun 2020 00:58:35 -0000

The concrete change here is twofold:

1. there are new limits (based on the most recent analysis of AES-GCM)

2. counts of (apparent) forgery attempts are not reset on key update

We contemplated doing the second for confidentiality limits as well, but=
 that was unlikely to preserve margins as making a new connection has th=
e same effect.

For me, the upshot was that - at least with current analyses - we don't =
have quite enough headroom in these ciphers.  We can't use them in all t=
he ways we might want AND keep very conservative safety margins.  The wo=
rst-case margins aren't *awful*, but they aren't in the realm of near im=
possibility that I like.

We use up a lot of headroom in assumptions and approximations.  There ar=
e likely gains to be had from refinements to analysis techniques or the =
threat model.  And, as Chris says, we don't have a thorough analysis of =
some modes; if that works out the way AES-GCM did, we will be in very go=
od shape.  Or maybe we'll just get a new AEAD with lots more headroom.

I don't expect this to be the final word on the subject, but this repres=
ents our current understanding on the subject and what we believe is the=
 best response based on that understanding.

On Thu, Jun 25, 2020, at 09:47, Christopher Wood wrote:
> Hi folks,=20
>=20
> I'm following up here based on an issue and proposed PR I just filed=20=

> against against draft-ietf-quic-tls.
>=20
>    Issue #3788: New confidentiality and integrity limits=20
> [https://github..com/quicwg/base-drafts/issues/3788
>    PR: https://github.com/quicwg/base-drafts/pull/3789
>=20
> TL;DR: We propose changing per-key limits to per-connection limits,=20=

> with different margins.=20
>=20
> Ongoing analysis efforts with Jean Paul Degabriele, Felix G=C3=BCnther=
,=20
> Kenny Paterson, and Martin Thomson revealed that QUIC should consider=20=

> changing its confidentiality and integrity limits to account for=20
> multi-user security settings. As some background, the existing limits=20=

> in draft-ietf-quic-tls apply to a single key. They are derived from=20=

> several existing analyses in the literature, including [1], [2], [3],=20=

> and [4]. Concretely, they seek to bound an attacker=E2=80=99s advantag=
e in=20
> breaking the security guarantees of the AEAD algorithm for a specific=20=

> key in question. When a Key Update occurs, endpoints receive a new key=
,=20
> which resets the encryption and forgery attempt counters for this new=20=

> key.=20
>=20
> This analysis fails to account for settings in which there are=20
> multiple, independent users of an AEAD algorithm, each with their own=20=

> unique key. In these so-called =E2=80=9Cmulti-user=E2=80=9D settings, =
the attacker is=20
> allowed to perform some amount of offline work to help accelerate any=20=

> attack on the AEAD algorithm. However, rather than target a single=20
> user=E2=80=99s specific key, they are tasked with distinguishing all t=
raffic=20
> (under all keys) from an equal amount of random bits. In the public-ke=
y=20
> setting, Bellare et al. [5] prove that the success probability in=20
> attacking one of the many independent users (or keys) is bounded by th=
e=20
> success probability of attacking a single user (or key) multiplied by=20=

> the number of users present. (This result extends to the symmetric-key=
=20
> setting.)
>=20
> Why is this important? Each key used throughout the duration of a=20
> single connection is effectively a distinct user. This means that the=20=

> attacker=E2=80=99s probability of success against any key in a connect=
ion must=20
> be considered across all keys. Or, more precisely, the integrity limit=
=20
> counters MUST NOT reset across Key Updates. (Confidentiality limits=20=

> still apply for individual keys, rather than across Key Updates, since=
=20
> there=E2=80=99s no functional difference in the analyses between teari=
ng down a=20
> connection and performing an update. The analyses consider encryption=20=

> as a function of plaintext blocks protected. Thus, users that need to=20=

> send N blocks of data will send N blocks of data, whether it be in one=
=20
> connection with many keys, or many connections with a single key.)
>=20
> Fortunately, this may not be much of a problem in practice. Hoang et=20=

> al. [6] provide tight multi-user security bounds for AES-GCM with nonc=
e=20
> randomization (as is used by TLS 1.3 and QUIC), and those bounds permi=
t=20
> a higher amount than previously established for a single key. However,=
=20
> to the best of our knowledge, there are no multi-user security analyse=
s=20
> giving tighter bounds than the generic analyses of=20
> AEAD_CHACHA20_POLY1305 and AEAD_AES_128_CCM. Thus, for the time being,=
=20
> we are stuck with the single-user integrity limits spread across Key=20=

> Updates for these AEADs. Recognizing that this is a relatively new are=
a=20
> for research, we are aiming to be quite conservative in setting limits=
,=20
> though we do allow an attacker a greater advantage in this multi-user=20=

> attack than we would for a single connection. We may relax these limit=
s=20
> if and when future analyses demonstrate that it=E2=80=99s safe to do s=
o.
>=20
> Note that this is not an ideal outcome. A =E2=80=9Ctrue=E2=80=9D multi=
-user setting=20
> would consider all users, i.e., all keys in all connections used by al=
l=20
> QUIC clients and servers, simultaneously. However, endpoints cannot=20=

> obtain such a global view of the Internet, and thus cannot make=20
> realistic parameter choices for bounds based on the number of users.=20=

> The compromise struck in this PR is to consider a narrow view of=20
> multi-user security, i.e., one in which the only =E2=80=9Cusers=E2=80=9D=
 in a=20
> connection are those introduced by Key Update messages. (This is=20
> certainly a gap in the literature worth exploring in the future.) We=20=

> also recommend setting very strong targets for attacker advantage for =
a=20
> single key, which we estimate will ensure that an attacker still has=20=

> limited advantage in the multi-user setting.
>=20
> Felix, Martin, and I did our best to document the basis for the=20
> analysis in this issue (and corresponding PR) in this CFRG I-D:
>=20
>    https://github.com/chris-wood/draft-wood-cfrg-aead-limits
>=20
> We believe the foundations for this analysis are sound, though we are=20=

> of course happy to learn if we made arithmetic mistakes. :-)
>=20
> Thanks,
> Chris, on behalf of Jean Paul, Felix, Kenny, and Martin
>=20
> [1] https://eprint.iacr.org/2012/438.pdf
> [2] https://eprint.iacr.org/2014/613.pdf
> [3] https://link.springer.com/chapter/10.1007/3-540-36492-7_7
> [4] https://www.isg.rhul.ac.uk/~kp/TLS-AEbounds.pdf
> [5] https://cseweb.ucsd.edu/~mihir/papers/musu.pdf
> [6] https://dl.acm.org/doi/10.1145/3243734.3243816
>=20
>


From nobody Wed Jun 24 18:11:58 2020
Return-Path: <huitema@huitema.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 131BB3A1211 for <quic@ietfa.amsl.com>; Wed, 24 Jun 2020 18:11:57 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.899
X-Spam-Level: 
X-Spam-Status: No, score=-1.899 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id tRsEsbwjKGL7 for <quic@ietfa.amsl.com>; Wed, 24 Jun 2020 18:11:55 -0700 (PDT)
Received: from mx36-out10.antispamcloud.com (mx36-out10.antispamcloud.com [209.126.121.30]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 2F30A3A120E for <quic@ietf.org>; Wed, 24 Jun 2020 18:11:54 -0700 (PDT)
Received: from xse333.mail2web.com ([66.113.197.79] helo=xse.mail2web.com) by mx169.antispamcloud.com with esmtp (Exim 4.92) (envelope-from <huitema@huitema.net>) id 1joGQa-000sl7-G0 for quic@ietf.org; Thu, 25 Jun 2020 03:11:54 +0200
Received: from xsmtp21.mail2web.com (unknown [10.100.68.60]) by xse.mail2web.com (Postfix) with ESMTPS id 49shkx6KCTz2M46 for <quic@ietf.org>; Wed, 24 Jun 2020 18:10:01 -0700 (PDT)
Received: from [10.5.2.14] (helo=xmail04.myhosting.com) by xsmtp21.mail2web.com with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:256) (Exim 4.92) (envelope-from <huitema@huitema.net>) id 1joGOv-0003i7-Ok for quic@ietf.org; Wed, 24 Jun 2020 18:10:01 -0700
Received: (qmail 12876 invoked from network); 25 Jun 2020 01:10:01 -0000
Received: from unknown (HELO [192.168.1.107]) (Authenticated-user:_huitema@huitema.net@[172.58.46.130]) (envelope-sender <huitema@huitema.net>) by xmail04.myhosting.com (qmail-ldap-1.03) with ESMTPA for <quic@ietf.org>; 25 Jun 2020 01:10:01 -0000
To: Christopher Wood <caw@heapingbits.net>, quic@ietf.org
References: <43bb1525-0afa-4c4e-a234-f6ccfacbdd29@www.fastmail.com>
From: Christian Huitema <huitema@huitema.net>
Autocrypt: addr=huitema@huitema.net; prefer-encrypt=mutual; keydata= mDMEXtavGxYJKwYBBAHaRw8BAQdA1ou9A5MHTP9N3jfsWzlDZ+jPnQkusmc7sfLmWVz1Rmu0 J0NocmlzdGlhbiBIdWl0ZW1hIDxodWl0ZW1hQGh1aXRlbWEubmV0PoiWBBMWCAA+FiEEw3G4 Nwi4QEpAAXUUELAmqKBYtJQFAl7WrxsCGwMFCQlmAYAFCwkIBwIGFQoJCAsCBBYCAwECHgEC F4AACgkQELAmqKBYtJQbMwD/ebj/qnSbthC/5kD5DxZ/Ip0CGJw5QBz/+fJp3R8iAlsBAMjK r2tmyWyJz0CUkVG24WaR5EAJDvgwDv8h22U6QVkAuDgEXtavGxIKKwYBBAGXVQEFAQEHQJoM 6MUAIqpoqdCIiACiEynZf7nlJg2Eu0pXIhbUGONdAwEIB4h+BBgWCAAmFiEEw3G4Nwi4QEpA AXUUELAmqKBYtJQFAl7WrxsCGwwFCQlmAYAACgkQELAmqKBYtJRm2wD7BzeK5gEXSmBcBf0j BYdSaJcXNzx4yPLbP4GnUMAyl2cBAJzcsR4RkwO4dCRqM9CHpVJCwHtbUDJaa55//E0kp+gH
Subject: Re: New confidentiality and integrity limits
Message-ID: <a4d7e8cf-5879-6941-e059-5d2f3e67ae86@huitema.net>
Date: Wed, 24 Jun 2020 18:10:01 -0700
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:68.0) Gecko/20100101 Thunderbird/68.9.0
MIME-Version: 1.0
In-Reply-To: <43bb1525-0afa-4c4e-a234-f6ccfacbdd29@www.fastmail.com>
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Content-Language: en-US
X-Originating-IP: 66.113.197.79
X-Spampanel-Domain: xsmtpout.mail2web.com
X-Spampanel-Username: 66.113.197.0/24
Authentication-Results: antispamcloud.com; auth=pass smtp.auth=66.113.197.0/24@xsmtpout.mail2web.com
X-Spampanel-Outgoing-Class: unsure
X-Spampanel-Outgoing-Evidence: Combined (0.15)
X-Recommended-Action: accept
X-Filter-ID: Mvzo4OR0dZXEDF/gcnlw0f6LF1GdvkEexklpcFpSF5apSDasLI4SayDByyq9LIhVUZbR67CQ7/vm /hHDJU4RXkTNWdUk1Ol2OGx3IfrIJKywOmJyM1qr8uRnWBrbSAGDcnqpk5VeF3xR4kF6iVwRtbgN zB/4Jkrw1eDLcif59ftqd9MAh6+PE/scPcm/ACndU7Tmz6iKnkQL9gqsxD347235Nhqq+/HvroPq 8GSPg+60/QPNqXybIny9WGhadIo/1Ofn8DBBgDynfYj8uZOR2ryme9ldZJ7uNXfg/GfS8fUvP/L5 rCqHDsKZM+xa1iwJX+gRCHfMVnsAk591zk0uilUI+ZL4xWiN8NS6C+dmX6OEdA4u1aThyWrQ/ou2 +v/lmX4Em37yFgrCB6NHRn1g+f3uncIqYSL3lhh5c81YyJqFoLZMmkWsaurVZfvqROaDnDtHb8z5 dpPkEuJ8Snwqla7jUnW3hy14Yji8fo+4xCnSRo4Rcu5Z37rMuDjCny5fE9ykbJ7I9co1MAEE3ruN Xsm8UJsAPvDcVSKtDCYkioPY5Qx4fJOk03R5fJtf/Dv/dkIzS7m4GUpXCY1Y3j3ilQ8EW2aDkYIa 2M5pALviQra5n1qAU0VTFRks+evUvJt4mhP0mU7v5izoFH0HNZLbcgwWr7Zk/vFgNdJniksrUhAx iMxSpkvqIEtRL3s4ePxvne6Agjui5gKB/Byw/yqfyPKY2AXNZGS5G93aGyH8MqMlOQRMVMd0HCeT skOZ5TL8PQsBQED/m9MQs2avfela3TXg724gFzhHYUe+7aKm0vWOXKxN13EcU2ca0tDxf/UiTi+J 2sBvM/O0p+zizleC4lU8fDj1CnRx+r4b/1Q/PZ86zB2Vu6+TJiLath6PGP2/vOnCUbNPgcPcQwzM gKHyQxUo+ql2ySTkvEFH/23XMww2BnTTFGX5/yI4Ky+1ZJcbGqc5H4PEZHeoI/d6LWFf332z7LMw LGdoi9FMQ5j9dQUvMi1YKAun15JQSJLyCT5k+MTObVKxHy/dols381l9r9ft9daDonlwd6LnuX+J u10=
X-Report-Abuse-To: spam@quarantine11.antispamcloud.com
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/uxIaaNM2GJKdsMJWNaimwOYBMS4>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 25 Jun 2020 01:11:57 -0000

Chris,

I would like to understand how exactly we should quantify the
multi-users set. You write that "Each key used throughout the duration
of a single connection is effectively a distinct user. This means that
the attacker=E2=80=99s probability of success against any key in a connec=
tion
must be considered across all keys." OK, but what about keys used in
different connections, such as multiple clients connecting to the same
server, or for that matter multiple clients connecting to different
servers? Is this because all the keys used in the same connection are
derived from the same master secret?

If derivation from the same master secret is the issue, that means that
the protection offered by the key update function is seriously limited.
Otherwise, the keys used at different epochs would effectively be
independent. Obviously, we cannot design a new update function that
would somehow incorporate more entropy, at least not for QUIC V1. Does
that mean that the update mechanism is not very useful? Is it insecure
for applications to maintain long connections and rely on key update?
Should application rely instead on splitting the workload onto several
QUIC connection, each with its own master secret?

-- Christian huitema

On 6/24/2020 4:47 PM, Christopher Wood wrote:
> Hi folks,=20
>
> I'm following up here based on an issue and proposed PR I just filed ag=
ainst against draft-ietf-quic-tls.
>
>    Issue #3788: New confidentiality and integrity limits [https://githu=
b..com/quicwg/base-drafts/issues/3788
>    PR: https://github.com/quicwg/base-drafts/pull/3789
>
> TL;DR: We propose changing per-key limits to per-connection limits, wit=
h different margins.=20
>
> Ongoing analysis efforts with Jean Paul Degabriele, Felix G=C3=BCnther,=
 Kenny Paterson, and Martin Thomson revealed that QUIC should consider ch=
anging its confidentiality and integrity limits to account for multi-user=
 security settings. As some background, the existing limits in draft-ietf=
-quic-tls apply to a single key. They are derived from several existing a=
nalyses in the literature, including [1], [2], [3], and [4]. Concretely, =
they seek to bound an attacker=E2=80=99s advantage in breaking the securi=
ty guarantees of the AEAD algorithm for a specific key in question. When =
a Key Update occurs, endpoints receive a new key, which resets the encryp=
tion and forgery attempt counters for this new key.=20
>
> This analysis fails to account for settings in which there are multiple=
, independent users of an AEAD algorithm, each with their own unique key.=
 In these so-called =E2=80=9Cmulti-user=E2=80=9D settings, the attacker i=
s allowed to perform some amount of offline work to help accelerate any a=
ttack on the AEAD algorithm. However, rather than target a single user=E2=
=80=99s specific key, they are tasked with distinguishing all traffic (un=
der all keys) from an equal amount of random bits. In the public-key sett=
ing, Bellare et al. [5] prove that the success probability in attacking o=
ne of the many independent users (or keys) is bounded by the success prob=
ability of attacking a single user (or key) multiplied by the number of u=
sers present. (This result extends to the symmetric-key setting.)
>
> Why is this important? Each key used throughout the duration of a singl=
e connection is effectively a distinct user. This means that the attacker=
=E2=80=99s probability of success against any key in a connection must be=
 considered across all keys. Or, more precisely, the integrity limit coun=
ters MUST NOT reset across Key Updates. (Confidentiality limits still app=
ly for individual keys, rather than across Key Updates, since there=E2=80=
=99s no functional difference in the analyses between tearing down a conn=
ection and performing an update. The analyses consider encryption as a fu=
nction of plaintext blocks protected. Thus, users that need to send N blo=
cks of data will send N blocks of data, whether it be in one connection w=
ith many keys, or many connections with a single key.)
>
> Fortunately, this may not be much of a problem in practice. Hoang et al=
=2E [6] provide tight multi-user security bounds for AES-GCM with nonce r=
andomization (as is used by TLS 1.3 and QUIC), and those bounds permit a =
higher amount than previously established for a single key. However, to t=
he best of our knowledge, there are no multi-user security analyses givin=
g tighter bounds than the generic analyses of AEAD_CHACHA20_POLY1305 and =
AEAD_AES_128_CCM. Thus, for the time being, we are stuck with the single-=
user integrity limits spread across Key Updates for these AEADs. Recogniz=
ing that this is a relatively new area for research, we are aiming to be =
quite conservative in setting limits, though we do allow an attacker a gr=
eater advantage in this multi-user attack than we would for a single conn=
ection. We may relax these limits if and when future analyses demonstrate=
 that it=E2=80=99s safe to do so.
>
> Note that this is not an ideal outcome. A =E2=80=9Ctrue=E2=80=9D multi-=
user setting would consider all users, i.e., all keys in all connections =
used by all QUIC clients and servers, simultaneously. However, endpoints =
cannot obtain such a global view of the Internet, and thus cannot make re=
alistic parameter choices for bounds based on the number of users. The co=
mpromise struck in this PR is to consider a narrow view of multi-user sec=
urity, i.e., one in which the only =E2=80=9Cusers=E2=80=9D in a connectio=
n are those introduced by Key Update messages. (This is certainly a gap i=
n the literature worth exploring in the future.) We also recommend settin=
g very strong targets for attacker advantage for a single key, which we e=
stimate will ensure that an attacker still has limited advantage in the m=
ulti-user setting.
>
> Felix, Martin, and I did our best to document the basis for the analysi=
s in this issue (and corresponding PR) in this CFRG I-D:
>
>    https://github.com/chris-wood/draft-wood-cfrg-aead-limits
>
> We believe the foundations for this analysis are sound, though we are o=
f course happy to learn if we made arithmetic mistakes. :-)
>
> Thanks,
> Chris, on behalf of Jean Paul, Felix, Kenny, and Martin
>
> [1] https://eprint.iacr.org/2012/438.pdf
> [2] https://eprint.iacr.org/2014/613.pdf
> [3] https://link.springer.com/chapter/10.1007/3-540-36492-7_7
> [4] https://www.isg.rhul.ac.uk/~kp/TLS-AEbounds.pdf
> [5] https://cseweb.ucsd.edu/~mihir/papers/musu.pdf
> [6] https://dl.acm.org/doi/10.1145/3243734.3243816
>


From nobody Thu Jun 25 11:29:45 2020
Return-Path: <caw@heapingbits.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 4404C3A0F87 for <quic@ietfa.amsl.com>; Thu, 25 Jun 2020 11:29:42 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_BLOCKED=0.001, RCVD_IN_MSPIKE_H3=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=heapingbits.net header.b=faGApYsh; dkim=pass (2048-bit key) header.d=messagingengine.com header.b=jKmPimP+
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id wwsZWHn1bGaQ for <quic@ietfa.amsl.com>; Thu, 25 Jun 2020 11:29:39 -0700 (PDT)
Received: from out2-smtp.messagingengine.com (out2-smtp.messagingengine.com [66.111.4.26]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 6F0EA3A0F67 for <quic@ietf.org>; Thu, 25 Jun 2020 11:29:34 -0700 (PDT)
Received: from compute1.internal (compute1.nyi.internal [10.202.2.41]) by mailout.nyi.internal (Postfix) with ESMTP id B750E5C0159; Thu, 25 Jun 2020 14:29:33 -0400 (EDT)
Received: from imap4 ([10.202.2.54]) by compute1.internal (MEProxy); Thu, 25 Jun 2020 14:29:33 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=heapingbits.net; h=mime-version:message-id:in-reply-to:references:date:from:to :subject:content-type:content-transfer-encoding; s=fm1; bh=P4TmE sl1Mo2CA16Z7wzSFuo+tRioQhl2SovQLAtNXB4=; b=faGApYshCXddrUlYSX6W0 bajtfKRFfLHEPePfvEZ0fBY5L4mmc4PJtEXru/93qsauZ6itV6QArm1eA/daM0RI 47nJ/Y3K0yNUnP7EQLghstb0WVojDhs37+kom6L39CLvBbF30xF4RGcTMHVM85dQ 96J/xFQzS50LzVtbifh1hrxVxVXYng22rCnUNR8UwcTuCTJC27+CzY2IYGv13VvV yXwoEr2+V/JpAy0mCIMY1BC8HcFPmJFQdn1VkeGr7tGNRl7Js0Bkpg0gGfwpZY+m QAN2DC9iWwxlNmsSEAMFodCcyEljpJnrKjpVGvP8EJRtJYttaP69te85/LB/kylu A==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d= messagingengine.com; h=content-transfer-encoding:content-type :date:from:in-reply-to:message-id:mime-version:references :subject:to:x-me-proxy:x-me-proxy:x-me-sender:x-me-sender :x-sasl-enc; s=fm3; bh=P4TmEsl1Mo2CA16Z7wzSFuo+tRioQhl2SovQLAtNX B4=; b=jKmPimP+3aQyPTHPmivrrFXzvcUxePVmWIvg8CMUUPvNy8ueLnOrlBHk5 30BvA5vi02B/ECdnz0ARcAPh9i8al4lL1ZGMXiFr5wDOCj4lID7FZg7Mz/PGEpJ3 qu2aMsbl0OLXPjTh/IdN58EGJghBYWXsgXwxsvwGaWlz6C3kKxCec1GPkEFWitqP MsEvNCL47xHiWodHATo02U93HOAA3tzfD+bJg0QlRcc8ucxUge2Fw/EHZDHuqho7 0v+FxlaQG+l+oTJo+oMvoMIQXridqHvgY4cemQ038Kf1gW+P1xslieaeahajJJyS v7/3fFMbG0uXcMsI5LeM7nJUZLPjA==
X-ME-Sender: <xms:De30XjsyiTGeQ0TZZPfw3EdxO8uOxjrNE0xjz_qlO_IHMUTL0JHGGA>
X-ME-Proxy-Cause: gggruggvucftvghtrhhoucdtuddrgeduhedrudekledguddvjecutefuodetggdotefrod ftvfcurfhrohhfihhlvgemucfhrghsthforghilhdpqfgfvfdpuffrtefokffrpgfnqfgh necuuegrihhlohhuthemuceftddtnecusecvtfgvtghiphhivghnthhsucdlqddutddtmd enucfjughrpefofgggkfgjfhffhffvufgtgfesthhqredtreerjeenucfhrhhomhepfdev hhhrihhsthhophhhvghrucghohhougdfuceotggrfieshhgvrghpihhnghgsihhtshdrnh gvtheqnecuggftrfgrthhtvghrnheptdeiveeivdevieeuleelgfekfefgffffudduvefg gfffleekgfduuefggfehgfdtnecuffhomhgrihhnpehgihhthhhusgdrrdgtohhmpdhgih hthhhusgdrtghomhdpihgrtghrrdhorhhgpdhsphhrihhnghgvrhdrtghomhdprhhhuhhl rdgrtgdruhhkpdhutghsugdrvgguuhdprggtmhdrohhrghenucevlhhushhtvghrufhiii gvpedtnecurfgrrhgrmhepmhgrihhlfhhrohhmpegtrgifsehhvggrphhinhhgsghithhs rdhnvght
X-ME-Proxy: <xmx:De30XkeB5RrUgSdShjp85oJG8WYOt21BNsg_9lrXByR9gtVA4r6eIw> <xmx:De30XmwtslzI3z3_3-4QaDvZu2oagMoBzV52sk8jV95l_0vhVDMvdw> <xmx:De30XiPLH1lQc37Eg9jHcspCYLabKiXNQ1UL46fV4r_eQjRS5VIpdg> <xmx:De30XmKJuQxzsdW3gaSbliwN2F4tiR8rEkKPUjxYzyElSmS2EjYSqQ>
Received: by mailuser.nyi.internal (Postfix, from userid 501) id 423393C00A1; Thu, 25 Jun 2020 14:29:33 -0400 (EDT)
X-Mailer: MessagingEngine.com Webmail Interface
User-Agent: Cyrus-JMAP/3.3.0-dev0-543-gda70334-fm-20200618.004-gda703345
Mime-Version: 1.0
Message-Id: <58b40fc0-86ad-49e9-bc17-187c38e97f7a@www.fastmail.com>
In-Reply-To: <a4d7e8cf-5879-6941-e059-5d2f3e67ae86@huitema.net>
References: <43bb1525-0afa-4c4e-a234-f6ccfacbdd29@www.fastmail.com> <a4d7e8cf-5879-6941-e059-5d2f3e67ae86@huitema.net>
Date: Thu, 25 Jun 2020 11:29:13 -0700
From: "Christopher Wood" <caw@heapingbits.net>
To: "Christian Huitema" <huitema@huitema.net>, quic@ietf.org
Subject: Re: New confidentiality and integrity limits
Content-Type: text/plain;charset=utf-8
Content-Transfer-Encoding: quoted-printable
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/xem4LOPR5nrleKtEvt3-gEqjWLw>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 25 Jun 2020 18:29:43 -0000

Hi Christian,

On Wed, Jun 24, 2020, at 6:10 PM, Christian Huitema wrote:
> I would like to understand how exactly we should quantify the
> multi-users set. You write that "Each key used throughout the duration=

> of a single connection is effectively a distinct user. This means that=

> the attacker=E2=80=99s probability of success against any key in a con=
nection
> must be considered across all keys." OK, but what about keys used in
> different connections, such as multiple clients connecting to the same=

> server, or for that matter multiple clients connecting to different
> servers? Is this because all the keys used in the same connection are
> derived from the same master secret?

In the multi-user setting, a user is effectively a distinct key. Thus, t=
he set of users is all keys used across all connections. (Key updates gi=
ve what is effectively a new independent key, so they introduce new "use=
rs" into the system.)

> If derivation from the same master secret is the issue, that means tha=
t
> the protection offered by the key update function is seriously limited=
.
> Otherwise, the keys used at different epochs would effectively be
> independent. Obviously, we cannot design a new update function that
> would somehow incorporate more entropy, at least not for QUIC V1. Does=

> that mean that the update mechanism is not very useful? Is it insecure=

> for applications to maintain long connections and rely on key update?
> Should application rely instead on splitting the workload onto several=

> QUIC connection, each with its own master secret?

For confidentiality, the important bit is to keep the amount of data enc=
rypted per key "reasonably low." Running a key update before the limits =
are reached helps drive down the adversary's advantage. So, it's fine to=
 maintain a long-lived connection provided the confidentiality limits ar=
e not violated and endpoints update frequently.=20

As Martin noted, for the integrity limit, the story is slightly differen=
t. Key updates do not help there.

Best,
Chris

>=20
> -- Christian huitema
>=20
> On 6/24/2020 4:47 PM, Christopher Wood wrote:
> > Hi folks,=20
> >
> > I'm following up here based on an issue and proposed PR I just filed=
 against against draft-ietf-quic-tls.
> >
> >    Issue #3788: New confidentiality and integrity limits [https://gi=
thub..com/quicwg/base-drafts/issues/3788
> >    PR: https://github.com/quicwg/base-drafts/pull/3789
> >
> > TL;DR: We propose changing per-key limits to per-connection limits, =
with different margins.=20
> >
> > Ongoing analysis efforts with Jean Paul Degabriele, Felix G=C3=BCnth=
er, Kenny Paterson, and Martin Thomson revealed that QUIC should conside=
r changing its confidentiality and integrity limits to account for multi=
-user security settings. As some background, the existing limits in draf=
t-ietf-quic-tls apply to a single key. They are derived from several exi=
sting analyses in the literature, including [1], [2], [3], and [4]. Conc=
retely, they seek to bound an attacker=E2=80=99s advantage in breaking t=
he security guarantees of the AEAD algorithm for a specific key in quest=
ion. When a Key Update occurs, endpoints receive a new key, which resets=
 the encryption and forgery attempt counters for this new key.=20
> >
> > This analysis fails to account for settings in which there are multi=
ple, independent users of an AEAD algorithm, each with their own unique =
key. In these so-called =E2=80=9Cmulti-user=E2=80=9D settings, the attac=
ker is allowed to perform some amount of offline work to help accelerate=
 any attack on the AEAD algorithm. However, rather than target a single =
user=E2=80=99s specific key, they are tasked with distinguishing all tra=
ffic (under all keys) from an equal amount of random bits. In the public=
-key setting, Bellare et al. [5] prove that the success probability in a=
ttacking one of the many independent users (or keys) is bounded by the s=
uccess probability of attacking a single user (or key) multiplied by the=
 number of users present. (This result extends to the symmetric-key sett=
ing.)
> >
> > Why is this important? Each key used throughout the duration of a si=
ngle connection is effectively a distinct user. This means that the atta=
cker=E2=80=99s probability of success against any key in a connection mu=
st be considered across all keys. Or, more precisely, the integrity limi=
t counters MUST NOT reset across Key Updates. (Confidentiality limits st=
ill apply for individual keys, rather than across Key Updates, since the=
re=E2=80=99s no functional difference in the analyses between tearing do=
wn a connection and performing an update. The analyses consider encrypti=
on as a function of plaintext blocks protected. Thus, users that need to=
 send N blocks of data will send N blocks of data, whether it be in one =
connection with many keys, or many connections with a single key.)
> >
> > Fortunately, this may not be much of a problem in practice. Hoang et=
 al. [6] provide tight multi-user security bounds for AES-GCM with nonce=
 randomization (as is used by TLS 1.3 and QUIC), and those bounds permit=
 a higher amount than previously established for a single key. However, =
to the best of our knowledge, there are no multi-user security analyses =
giving tighter bounds than the generic analyses of AEAD_CHACHA20_POLY130=
5 and AEAD_AES_128_CCM. Thus, for the time being, we are stuck with the =
single-user integrity limits spread across Key Updates for these AEADs. =
Recognizing that this is a relatively new area for research, we are aimi=
ng to be quite conservative in setting limits, though we do allow an att=
acker a greater advantage in this multi-user attack than we would for a =
single connection. We may relax these limits if and when future analyses=
 demonstrate that it=E2=80=99s safe to do so.
> >
> > Note that this is not an ideal outcome. A =E2=80=9Ctrue=E2=80=9D mul=
ti-user setting would consider all users, i.e., all keys in all connecti=
ons used by all QUIC clients and servers, simultaneously. However, endpo=
ints cannot obtain such a global view of the Internet, and thus cannot m=
ake realistic parameter choices for bounds based on the number of users.=
 The compromise struck in this PR is to consider a narrow view of multi-=
user security, i.e., one in which the only =E2=80=9Cusers=E2=80=9D in a =
connection are those introduced by Key Update messages. (This is certain=
ly a gap in the literature worth exploring in the future.) We also recom=
mend setting very strong targets for attacker advantage for a single key=
, which we estimate will ensure that an attacker still has limited advan=
tage in the multi-user setting.
> >
> > Felix, Martin, and I did our best to document the basis for the anal=
ysis in this issue (and corresponding PR) in this CFRG I-D:
> >
> >    https://github.com/chris-wood/draft-wood-cfrg-aead-limits
> >
> > We believe the foundations for this analysis are sound, though we ar=
e of course happy to learn if we made arithmetic mistakes. :-)
> >
> > Thanks,
> > Chris, on behalf of Jean Paul, Felix, Kenny, and Martin
> >
> > [1] https://eprint.iacr.org/2012/438.pdf
> > [2] https://eprint.iacr.org/2014/613.pdf
> > [3] https://link.springer.com/chapter/10.1007/3-540-36492-7_7
> > [4] https://www.isg.rhul.ac.uk/~kp/TLS-AEbounds.pdf
> > [5] https://cseweb.ucsd.edu/~mihir/papers/musu.pdf
> > [6] https://dl.acm.org/doi/10.1145/3243734.3243816
> >
>=20
>


From nobody Thu Jun 25 16:45:19 2020
Return-Path: <mt@lowentropy.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id AF7193A107B for <quic@ietfa.amsl.com>; Thu, 25 Jun 2020 16:45:17 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.098
X-Spam-Level: 
X-Spam-Status: No, score=-2.098 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_MSPIKE_H4=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=lowentropy.net header.b=oMh0XLJV; dkim=pass (2048-bit key) header.d=messagingengine.com header.b=gEwg3tin
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id DyVnHqd-eMtx for <quic@ietfa.amsl.com>; Thu, 25 Jun 2020 16:45:15 -0700 (PDT)
Received: from wout1-smtp.messagingengine.com (wout1-smtp.messagingengine.com [64.147.123.24]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id D8C613A1053 for <quic@ietf.org>; Thu, 25 Jun 2020 16:45:15 -0700 (PDT)
Received: from compute2.internal (compute2.nyi.internal [10.202.2.42]) by mailout.west.internal (Postfix) with ESMTP id 17105830 for <quic@ietf.org>; Thu, 25 Jun 2020 19:45:15 -0400 (EDT)
Received: from imap2 ([10.202.2.52]) by compute2.internal (MEProxy); Thu, 25 Jun 2020 19:45:15 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=lowentropy.net; h=mime-version:message-id:in-reply-to:references:date:from:to :subject:content-type; s=fm2; bh=6Fjl5PFOdfrFQSGF25brJhGie0nZ8Vh 0fi5mL1fq2eA=; b=oMh0XLJVlfqRR8BJlDP5c3lx2kGUFT0TXnI39dqi//RbCTL IENnrnloGm1mnl68/2kCcRnhkV/mEkG6TP8ibODWpELc3NQ8he3CkyaADecRGucO TUTn3/1WL6L7htXKV2txcPy4sMw/2VQxWU7Am9KXA/iaNFX6U5W4NJZVgPZxienm Yjb5nrNfB9K5rpZXxwz8I28MEsLOoBEi2SEGJ81nS4JPRv7UYL/XHVon3+bi9VN9 R+AK/QQKemqvFqzV+/jT0m4s3xQzFsVvifo01ppSOEOGf8NxOltnXhm+0okZoRmg hKWHIjq6D8Q9P28OAAp725x+JPq8Bypwy00xMHA==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d= messagingengine.com; h=content-type:date:from:in-reply-to :message-id:mime-version:references:subject:to:x-me-proxy :x-me-proxy:x-me-sender:x-me-sender:x-sasl-enc; s=fm3; bh=6Fjl5P FOdfrFQSGF25brJhGie0nZ8Vh0fi5mL1fq2eA=; b=gEwg3tinDPVXBHsJ35XsSg nCwDAjk8H4IJcD3pDpGYoqTehk8x+6KKjn55EZOZkedF9Lf6z+jrB+5K2Y4wNQ1Q sR6JoULb1CppKxkD+a1Lk2qgiaPR6OVkYGQfiWYppoDom5Pi4EVetSQ/NGk6IBl9 AzHWKugg6vh78QK+BJnRknXjQ1IygVvw/VNT0ijjedfV0pPit4YY/nGwl5cRX1ya CAWsTRK71ZNuHu4lryS7JRPHkPZhhbjqyyTxKhBDRutz/1crDm2gNlNiE3c+8WB3 Ya6KoPgyADgdiSNeYYcMM77KTfclf61Pz6ck0CYBpHWfkBGIQbzYBhi5VRKOESog ==
X-ME-Sender: <xms:Cjf1XnO0tBF91zDUwYmoK4IxMPYQ5G0We-r73jOPqTYPPWs-AyaCkQ>
X-ME-Proxy-Cause: gggruggvucftvghtrhhoucdtuddrgeduhedrudeltddgvdegucetufdoteggodetrfdotf fvucfrrhhofhhilhgvmecuhfgrshhtofgrihhlpdfqfgfvpdfurfetoffkrfgpnffqhgen uceurghilhhouhhtmecufedttdenucenucfjughrpefofgggkfgjfhffhffvufgtsehttd ertderredtnecuhfhrohhmpedfofgrrhhtihhnucfvhhhomhhsohhnfdcuoehmtheslhho figvnhhtrhhophihrdhnvghtqeenucggtffrrghtthgvrhhnpeekteeuieektdekleefke evhfekffevvdevgfekgfeluefgvdejjeegffeigedtjeenucevlhhushhtvghrufhiiigv pedtnecurfgrrhgrmhepmhgrihhlfhhrohhmpehmtheslhhofigvnhhtrhhophihrdhnvg ht
X-ME-Proxy: <xmx:Cjf1Xh9DCkC9A7bYf6dIegqQIR5r4jlhw2gP5rSYEKQs1H1RR2VVWw> <xmx:Cjf1XmTGoDZoyMXjUHekv2MBohpuCWPx5oxk4I-TpHAndzhrdu3-5A> <xmx:Cjf1Xru8xZWDtjsRXdmhDtvdgECZ84uE6U_sP1kQOBhKl9W1ul3emQ> <xmx:Cjf1Xi-Ye7uqV54K-nIyNNudHaSOa_PDjK11DmTM5hmGkQoNp7SASQ>
Received: by mailuser.nyi.internal (Postfix, from userid 501) id 488A8E00A8; Thu, 25 Jun 2020 19:45:14 -0400 (EDT)
X-Mailer: MessagingEngine.com Webmail Interface
User-Agent: Cyrus-JMAP/3.3.0-dev0-543-gda70334-fm-20200618.004-gda703345
Mime-Version: 1.0
Message-Id: <1d543593-7abc-4a89-93c5-82c464492786@www.fastmail.com>
In-Reply-To: <58b40fc0-86ad-49e9-bc17-187c38e97f7a@www.fastmail.com>
References: <43bb1525-0afa-4c4e-a234-f6ccfacbdd29@www.fastmail.com> <a4d7e8cf-5879-6941-e059-5d2f3e67ae86@huitema.net> <58b40fc0-86ad-49e9-bc17-187c38e97f7a@www.fastmail.com>
Date: Fri, 26 Jun 2020 09:44:55 +1000
From: "Martin Thomson" <mt@lowentropy.net>
To: quic@ietf.org
Subject: Re: New confidentiality and integrity limits
Content-Type: text/plain
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/e53V1ZDhk_dlnaEa4w_Qs8cNk8Y>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 25 Jun 2020 23:45:18 -0000

On Fri, Jun 26, 2020, at 04:29, Christopher Wood wrote:
> As Martin noted, for the integrity limit, the story is slightly 
> different. Key updates do not help there.

Or we don't know that they do...  There are still a lot of unanswered questions.  We do know that a forgery attempt against one key is likely still good for a different key on the same connection.  It might not be good for another connection (the inclusion of connection IDs might help us here, but that's another unknown - and unreliable - factor) but we can't really control for that.  So we chose something we believe to be both conservative and practical.

The forgery limits are still quite generous: if you have to drop a connection due to that many forgeries, then I'd suggest that something is seriously wrong.


From nobody Thu Jun 25 18:01:20 2020
Return-Path: <jri.ietf@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C203C3A10B7 for <quic@ietfa.amsl.com>; Thu, 25 Jun 2020 18:01:18 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id fdxgZNhRCFNL for <quic@ietfa.amsl.com>; Thu, 25 Jun 2020 18:01:17 -0700 (PDT)
Received: from mail-lj1-x22c.google.com (mail-lj1-x22c.google.com [IPv6:2a00:1450:4864:20::22c]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 4201F3A1046 for <quic@ietf.org>; Thu, 25 Jun 2020 18:01:17 -0700 (PDT)
Received: by mail-lj1-x22c.google.com with SMTP id h19so8512317ljg.13 for <quic@ietf.org>; Thu, 25 Jun 2020 18:01:17 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=cXONsrlIZNQNgHHOK4EepstCxg34Yc12VfYPR0r5HX4=; b=G4xBHZRl/lOMyuPJOF89qyf0XHB0QURgjFxHDrWC/JSkZgrwRTaXvL9apdupPWzyvM lfbncbLzxHlk1zMciDmJJ052RSkr6K0y5RRBBc9ylUwNIUqii+uYvn705gmAW99UhDeE svit4sTO6izbqSP6eRpvweso3WGuZFZZc42EjPdahqxQYLRLg2MD9JSovzvwYOuD58Z2 +lMeFoZQh4J77fUro95kU2nFJYyHbnkVXxml/r/ISvoigK6uc105HuegWEdqanMrTsBX TUgxwXMY66P4wb20Kzpmdu93+QSLLWLC8gGGLwJgw7jgW+bUafi1vum4DPnx57GtUHa/ WAyg==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=cXONsrlIZNQNgHHOK4EepstCxg34Yc12VfYPR0r5HX4=; b=m12Up12EkEFvxcaJjZxitKq5YfgO3/zUlYvW8mMGdGdg59ubzpOcS97U+/agDx5Crm H493VYYFqnegLeevUZH9BA0Zpae1ZJS2WGaTexaWoFDPGf3dB7468Cdp5jQOf2iVkA2S nN120gDvW1QjrbejVqi74JXBH3otKZe4wkm1oTneOTC3sAl5/f6GL+CejrLPJJXBcy1C 5mIgPf+eznP0ngNgizxlCtlyWmJlfco78HCagHFKbAc+SC3+70BXm2H/n5BjFp4k7Jcb HBBc7BQHSaF66wZrzGppOGP+UDlGO1+ivKowqb6ZyTEhLJfptifCGEROhVsoq6leov6o d0WA==
X-Gm-Message-State: AOAM532USCVlEykYBJ9jUFf4LSS0w9ioJzGFyWZDPQicL+IJmCtqpTkO t7wy69PeYjWykDjAuHA11A8u9PZjJ4k+TClB0Yc=
X-Google-Smtp-Source: ABdhPJx0272g5QahSgqO6YlQ0Kc8xQIouBMz8tBnkghXApmUUkMcj8mKP9PXQx5UL/5gxvbZDZYxuuGkzPHmYGPdDAA=
X-Received: by 2002:a2e:860d:: with SMTP id a13mr190563lji.344.1593133275229;  Thu, 25 Jun 2020 18:01:15 -0700 (PDT)
MIME-Version: 1.0
References: <43bb1525-0afa-4c4e-a234-f6ccfacbdd29@www.fastmail.com> <a4d7e8cf-5879-6941-e059-5d2f3e67ae86@huitema.net> <58b40fc0-86ad-49e9-bc17-187c38e97f7a@www.fastmail.com> <1d543593-7abc-4a89-93c5-82c464492786@www.fastmail.com>
In-Reply-To: <1d543593-7abc-4a89-93c5-82c464492786@www.fastmail.com>
From: Jana Iyengar <jri.ietf@gmail.com>
Date: Thu, 25 Jun 2020 18:01:04 -0700
Message-ID: <CACpbDccSB6H9AhmAinia7ojEg1VY-S=dF214wHvZcZkoiL-xFA@mail.gmail.com>
Subject: Re: New confidentiality and integrity limits
To: Martin Thomson <mt@lowentropy.net>
Cc: QUIC WG <quic@ietf.org>
Content-Type: multipart/alternative; boundary="000000000000f3336905a8f23e25"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/2rMVcbLT2KnKUlqHqEpa0h3b9B4>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 26 Jun 2020 01:01:19 -0000

--000000000000f3336905a8f23e25
Content-Type: text/plain; charset="UTF-8"

Chris, (and all others involved),

Thank you for all the analysis and for distilling it into the
recommendations above -- the more precise bounds are super helpful!
It seems to me that the other limits are not practical limits for the most
part. Am I correct in understanding that the practical impact of this boils
down to "We might need to do key updates more frequently than we had hoped"
since the integrity limits are a bit more modest?

I've left some comments on the PR.

Thanks again,
- jana

On Thu, Jun 25, 2020 at 4:45 PM Martin Thomson <mt@lowentropy.net> wrote:

> On Fri, Jun 26, 2020, at 04:29, Christopher Wood wrote:
> > As Martin noted, for the integrity limit, the story is slightly
> > different. Key updates do not help there.
>
> Or we don't know that they do...  There are still a lot of unanswered
> questions.  We do know that a forgery attempt against one key is likely
> still good for a different key on the same connection.  It might not be
> good for another connection (the inclusion of connection IDs might help us
> here, but that's another unknown - and unreliable - factor) but we can't
> really control for that.  So we chose something we believe to be both
> conservative and practical.
>
> The forgery limits are still quite generous: if you have to drop a
> connection due to that many forgeries, then I'd suggest that something is
> seriously wrong.
>
>

--000000000000f3336905a8f23e25
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div>Chris, (and all others involved),</div><div><br></div=
>Thank you for all the analysis and for distilling it into the recommendati=
ons above -- the more precise bounds are super helpful!<br>It seems to me t=
hat the other limits are not practical=C2=A0limits for the most part. Am I =
correct in understanding that the practical impact of this boils down to &q=
uot;We might need to do key updates more frequently than we had hoped&quot;=
 since the integrity limits are a bit more modest?<br><div><br></div><div>I=
&#39;ve left some comments on the PR.</div><div></div><div><br></div><div>T=
hanks again,</div><div>- jana</div></div><br><div class=3D"gmail_quote"><di=
v dir=3D"ltr" class=3D"gmail_attr">On Thu, Jun 25, 2020 at 4:45 PM Martin T=
homson &lt;<a href=3D"mailto:mt@lowentropy.net">mt@lowentropy.net</a>&gt; w=
rote:<br></div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0p=
x 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">On Fri, Ju=
n 26, 2020, at 04:29, Christopher Wood wrote:<br>
&gt; As Martin noted, for the integrity limit, the story is slightly <br>
&gt; different. Key updates do not help there.<br>
<br>
Or we don&#39;t know that they do...=C2=A0 There are still a lot of unanswe=
red questions.=C2=A0 We do know that a forgery attempt against one key is l=
ikely still good for a different key on the same connection.=C2=A0 It might=
 not be good for another connection (the inclusion of connection IDs might =
help us here, but that&#39;s another unknown - and unreliable - factor) but=
 we can&#39;t really control for that.=C2=A0 So we chose something we belie=
ve to be both conservative and practical.<br>
<br>
The forgery limits are still quite generous: if you have to drop a connecti=
on due to that many forgeries, then I&#39;d suggest that something is serio=
usly wrong.<br>
<br>
</blockquote></div>

--000000000000f3336905a8f23e25--


From nobody Thu Jun 25 18:48:54 2020
Return-Path: <jri.ietf@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id CC9823A10DD for <quic@ietfa.amsl.com>; Thu, 25 Jun 2020 18:48:51 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ocICYeN8x37V for <quic@ietfa.amsl.com>; Thu, 25 Jun 2020 18:48:50 -0700 (PDT)
Received: from mail-lf1-x131.google.com (mail-lf1-x131.google.com [IPv6:2a00:1450:4864:20::131]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 184A33A10DA for <quic@ietf.org>; Thu, 25 Jun 2020 18:48:49 -0700 (PDT)
Received: by mail-lf1-x131.google.com with SMTP id y13so4273040lfe.9 for <quic@ietf.org>; Thu, 25 Jun 2020 18:48:49 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=fRGoQcDpydBNzwiZGOpKlbUfQ2b4D69yLHCBZbpPIn4=; b=Dw8t604w8izeZ7K/gI9Bo38g6SnsoSP5ekPf/53yyVKwjDo0JFxGdrRStMlKhRqjVF kcLQMSlXO3zJgwUonnRGnQv/lNJtBhuiDj6/5oQRChnBRqT6ckn9yNnsHW9ECh+iNMES ncEhqUozAJgJz2+Q4E0B4sDWC9UJ2jJULdnJt4P/UwtmP1swZL7wTTlFWfsvvN39olsx mqc+ggD3XoTP9FqVDZsA9riWLwtR/2F51F9axd3LMQLECi2w83xNm16OEeezEK2tyq9s TBridXWvdBtoJnand2WJqIIdGeJiIz73JONLn8JxScPPB59YojRYTua60sHXwomEXCS3 SzkA==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=fRGoQcDpydBNzwiZGOpKlbUfQ2b4D69yLHCBZbpPIn4=; b=BfJPXEFs3RrX5ZuaifzU0ye1RzGBcZ+xExyp7X058+L5Q8zTqzt8USrwAaOUVGGNA1 1Ir1KbAmAo5VvkSCpkaQm+n9V60jDHa4QLCw2b+TysPKWP1nt9lInujdN4zIVd8g0NeT CfVZqpDFuFQDaS0MGuFKHCE/LWKtQQ88/jLi1s4Gsdi7E08kbzX9OZmiXULgZJJFao6y Pj2JYx8jUDGjaGloDc8HgVkbLXuWfg56T0H1hra7fURXkmVTN0whkGCxR0VvEuLkAkBc raxgh49XvzpZEqWMwsOwv9pmJph4sVSebken6xZk5agQHVzNXwgV57yqy1ANIk1WyW+p UhWA==
X-Gm-Message-State: AOAM530PDFATV8GP0sxRgw3IWOWloU0SfLcuhdqzD2eg3JciLQKJzzbs 0J3dBKUMXJQbToLw0UVM3XlCnq1gtxZeUI2TT0I=
X-Google-Smtp-Source: ABdhPJwpklJg4hbmJ2ouJyvjzKitzk3isA4JIV6WBjiY+x6VUnS8IOgkyHnWM9O4M2I67tJ+sgRUsFvBlDi+y1y/lcc=
X-Received: by 2002:ac2:4550:: with SMTP id j16mr456121lfm.37.1593136128238; Thu, 25 Jun 2020 18:48:48 -0700 (PDT)
MIME-Version: 1.0
References: <8c747b59-1672-4ccb-dfce-0914e74c504f@gmx.de> <9B975808-6BF3-491C-8E32-C475824B0E9A@mnot.net> <7DD3EE98-199B-4F09-A8C8-1CE604989136@gbiv.com>
In-Reply-To: <7DD3EE98-199B-4F09-A8C8-1CE604989136@gbiv.com>
From: Jana Iyengar <jri.ietf@gmail.com>
Date: Thu, 25 Jun 2020 18:48:37 -0700
Message-ID: <CACpbDcdCPMOqrSF_RgFy8NoVdFwNMocU5cOek6knp=Ux5gpN=w@mail.gmail.com>
Subject: Re: draft-ietf-quic-http-29, "1.1. Prior versions of HTTP"
To: "Roy T. Fielding" <fielding@gbiv.com>
Cc: Mark Nottingham <mnot@mnot.net>, Julian Reschke <julian.reschke@gmx.de>, QUIC WG <quic@ietf.org>
Content-Type: multipart/alternative; boundary="00000000000000a76905a8f2e958"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/ppZBrLlHSD1bzYKu61bUaM0OS8w>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 26 Jun 2020 01:48:52 -0000

--00000000000000a76905a8f2e958
Content-Type: text/plain; charset="UTF-8"

I've filed https://github.com/quicwg/base-drafts/issues/3790 for this
editorial issue.

On Fri, Jun 19, 2020 at 10:33 AM Roy T. Fielding <fielding@gbiv.com> wrote:

> > On Jun 18, 2020, at 4:19 PM, Mark Nottingham <mnot@mnot.net> wrote:
> >
> > My .02 -
> >
> > I think we need to balance between being correct and over-complicating
> the text here. In *theory*, Julian is correct; in wide practice, pipelining
> isn't used because of the limitations it has, as well as implementation /
> interop issues.
>
> No, it isn't used by browsers because they aren't aware of the purpose
> for each connection and request. API clients almost always use pipelines.
>
> > How about something like:
> >
> > "Because of limitations in how HTTP exchanges are mapped onto TCP
> connections in HTTP/1.1, multiple parallel connections are often used..."
>
> Why don't you just say
>
>   "Because HTTP/1.x does not include a multiplexing layer, multiple
>    TCP connections are often used to service requests in parallel.
>    However, that has a negative impact on congestion control and network
>    efficiency, since TCP does not share congestion control across multiple
>    connections."
>
> because those are the actual technical limitations.
>
> .....Roy
>
>

--00000000000000a76905a8f2e958
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">I&#39;ve filed=C2=A0<a href=3D"https://github.com/quicwg/b=
ase-drafts/issues/3790">https://github.com/quicwg/base-drafts/issues/3790</=
a>=C2=A0for this editorial issue.</div><br><div class=3D"gmail_quote"><div =
dir=3D"ltr" class=3D"gmail_attr">On Fri, Jun 19, 2020 at 10:33 AM Roy T. Fi=
elding &lt;<a href=3D"mailto:fielding@gbiv.com">fielding@gbiv.com</a>&gt; w=
rote:<br></div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0p=
x 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">&gt; On Ju=
n 18, 2020, at 4:19 PM, Mark Nottingham &lt;<a href=3D"mailto:mnot@mnot.net=
" target=3D"_blank">mnot@mnot.net</a>&gt; wrote:<br>
&gt; <br>
&gt; My .02 -<br>
&gt; <br>
&gt; I think we need to balance between being correct and over-complicating=
 the text here. In *theory*, Julian is correct; in wide practice, pipelinin=
g isn&#39;t used because of the limitations it has, as well as implementati=
on / interop issues.<br>
<br>
No, it isn&#39;t used by browsers because they aren&#39;t aware of the purp=
ose<br>
for each connection and request. API clients almost always use pipelines.<b=
r>
<br>
&gt; How about something like:<br>
&gt; <br>
&gt; &quot;Because of limitations in how HTTP exchanges are mapped onto TCP=
 connections in HTTP/1.1, multiple parallel connections are often used...&q=
uot;<br>
<br>
Why don&#39;t you just say<br>
<br>
=C2=A0 &quot;Because HTTP/1.x does not include a multiplexing layer, multip=
le<br>
=C2=A0 =C2=A0TCP connections are often used to service requests in parallel=
.<br>
=C2=A0 =C2=A0However, that has a negative impact on congestion control and =
network<br>
=C2=A0 =C2=A0efficiency, since TCP does not share congestion control across=
 multiple<br>
=C2=A0 =C2=A0connections.&quot;<br>
<br>
because those are the actual technical limitations.<br>
<br>
.....Roy<br>
<br>
</blockquote></div>

--00000000000000a76905a8f2e958--


From nobody Thu Jun 25 21:31:31 2020
Return-Path: <mt@lowentropy.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 812D43A1119 for <quic@ietfa.amsl.com>; Thu, 25 Jun 2020 21:31:29 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.098
X-Spam-Level: 
X-Spam-Status: No, score=-2.098 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_MSPIKE_H4=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=lowentropy.net header.b=OKPL74Pa; dkim=pass (2048-bit key) header.d=messagingengine.com header.b=aedA+Osp
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Zz6LUY1fOSg3 for <quic@ietfa.amsl.com>; Thu, 25 Jun 2020 21:31:28 -0700 (PDT)
Received: from out3-smtp.messagingengine.com (out3-smtp.messagingengine.com [66.111.4.27]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 2A5C63A10F8 for <quic@ietf.org>; Thu, 25 Jun 2020 21:31:28 -0700 (PDT)
Received: from compute2.internal (compute2.nyi.internal [10.202.2.42]) by mailout.nyi.internal (Postfix) with ESMTP id 75F405C00B7; Fri, 26 Jun 2020 00:31:27 -0400 (EDT)
Received: from imap2 ([10.202.2.52]) by compute2.internal (MEProxy); Fri, 26 Jun 2020 00:31:27 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=lowentropy.net; h=mime-version:message-id:in-reply-to:references:date:from:to :cc:subject:content-type; s=fm2; bh=Fk7z/UZqqpofwozX5agIFBJeAiCA Zh3H8Tu7HQ3Xjb0=; b=OKPL74PaaKA0pWHWz5RwL5pxZWvrjspLExFeyFQ/KkWz 6h3hxsQI2WpqPZYfDU6ixHOWnx1juWR0Y4PNHfrFmdZVid6T/Mgo+BTI7yd04oPO 8FYisTcVngsStBrslSVIfxbi1Dd3Tj5DvuLBdDz6agv1gXvByKOmTPI5A5CCNwnv nczGTYEIYJfUA5EDjFp87h/dzs6j0yeT2iudJba2TtjURWK5wTjrlLG3sABt7XMH PurmieX0U1vZOwUb3TyjJkuJsJi10TYzqSfiSkmH/Yt6drYPmHUhP6gtNeBuW3WU bDEYYUvSX+Grdg4qe/0VAnrxCdMX4gYHjERO+R70fQ==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d= messagingengine.com; h=cc:content-type:date:from:in-reply-to :message-id:mime-version:references:subject:to:x-me-proxy :x-me-proxy:x-me-sender:x-me-sender:x-sasl-enc; s=fm3; bh=Fk7z/U ZqqpofwozX5agIFBJeAiCAZh3H8Tu7HQ3Xjb0=; b=aedA+OspWktuovPrB3Y2Oi 7nCUv4/3PEMPZUqtp/MRdjJ5C/dPoS0YA7MJBqnmF6d73YfYoVmIQS/PZYI1IkSQ oILV10DiJHWP0tW9Zqc+ss7HF7KB/R8ugB+dPRBf/SRB/hr0XNFO2JWTbEjeT5C8 nSvBxwrvIUR//qhJ+Z/1zI6TrSkIgNF7XpY6iki1x8Ty+tMBhX2bzUCK3bWwoL+7 Oz2FvTku/UrbO3TgeiCo7Gb21TY9Tk6hs2SFaGdWDlp48wfcueYPT4JiuVNlJYnj 4CBM6+8xL1xL3O/l0Uk+W85mdz30VtRA8tOWo1riHV0Nva+7x3C0dTx5xG0jvnjA ==
X-ME-Sender: <xms:H3r1XsfoCv1qkpxuKJuCoyqsh2y57Po6MWPxty7A-_WhOCkRwunmiA>
X-ME-Proxy-Cause: gggruggvucftvghtrhhoucdtuddrgeduhedrudeltddgkeefucetufdoteggodetrfdotf fvucfrrhhofhhilhgvmecuhfgrshhtofgrihhlpdfqfgfvpdfurfetoffkrfgpnffqhgen uceurghilhhouhhtmecufedttdenucesvcftvggtihhpihgvnhhtshculddquddttddmne cujfgurhepofgfggfkjghffffhvffutgesthdtredtreertdenucfhrhhomhepfdforghr thhinhcuvfhhohhmshhonhdfuceomhhtsehlohifvghnthhrohhphidrnhgvtheqnecugg ftrfgrthhtvghrnhepkeetueeikedtkeelfeekvefhkeffvedvvefgkefgleeugfdvjeej geffieegtdejnecuvehluhhsthgvrhfuihiivgeptdenucfrrghrrghmpehmrghilhhfrh homhepmhhtsehlohifvghnthhrohhphidrnhgvth
X-ME-Proxy: <xmx:H3r1XuMCMvzddAi7ivV2-b0fjVAT3tbzJzIEqX_DoBch8NQ_58nseg> <xmx:H3r1XtiEh8o7QsqHwspS0SvLRU1SAPQB4UZryRzN5SKiZA8igGZ4Cw> <xmx:H3r1Xh8MEQnCDhypH0KcFYWGlAHrtmqpLgu9Skeyd7zMrGB9D2jQFg> <xmx:H3r1Xm4FDy-3QAcoS8biaDfhlMX3l6li5mtUdNhR4NKAj4HxbesstA>
Received: by mailuser.nyi.internal (Postfix, from userid 501) id 299E1E00A8; Fri, 26 Jun 2020 00:31:27 -0400 (EDT)
X-Mailer: MessagingEngine.com Webmail Interface
User-Agent: Cyrus-JMAP/3.3.0-dev0-543-gda70334-fm-20200618.004-gda703345
Mime-Version: 1.0
Message-Id: <7d549475-249b-4674-9aa8-c2bf9953243b@www.fastmail.com>
In-Reply-To: <CACpbDccSB6H9AhmAinia7ojEg1VY-S=dF214wHvZcZkoiL-xFA@mail.gmail.com>
References: <43bb1525-0afa-4c4e-a234-f6ccfacbdd29@www.fastmail.com> <a4d7e8cf-5879-6941-e059-5d2f3e67ae86@huitema.net> <58b40fc0-86ad-49e9-bc17-187c38e97f7a@www.fastmail.com> <1d543593-7abc-4a89-93c5-82c464492786@www.fastmail.com> <CACpbDccSB6H9AhmAinia7ojEg1VY-S=dF214wHvZcZkoiL-xFA@mail.gmail.com>
Date: Fri, 26 Jun 2020 14:31:03 +1000
From: "Martin Thomson" <mt@lowentropy.net>
To: "Jana Iyengar" <jri.ietf@gmail.com>
Cc: "QUIC WG" <quic@ietf.org>
Subject: Re: New confidentiality and integrity limits
Content-Type: text/plain
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/exTpp0UrJ4N-B55F6BhtzoZDf2c>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 26 Jun 2020 04:31:30 -0000

On Fri, Jun 26, 2020, at 11:01, Jana Iyengar wrote:
> Thank you for all the analysis and for distilling it into the 
> recommendations above -- the more precise bounds are super helpful!
> It seems to me that the other limits are not practical limits for the 
> most part. 

Assuming that you mean "the limits are unlikely to be a problem in practice"?  I think that is right.

The numbers that change here are forgery limits, and we expect that a real connection will have very few apparent forgeries.  In testing, I see a handful during the handshake (with my RTT, I often get Initial retransmissions that arrive after the keys have been dropped), but that's only ever about 2 packets from what I've seen.  Similarly, an implementation that drops keys aggressively around a key update might see a few near each update.  

Overall, we're talking about seeing a handful of apparent forgeries in normal use, not millions (which is the lowest limit we have).  This could mean that someone using AES-CCM will run into this limit before they hit the 2^62 packet limit, but that is unlikely to be problem for the other ciphers as their limits are higher.

This is mostly a problem for people who might be subject to sustained forgery attacks, but an attacker that sends ~8 million bad packets to close a connection has earned their ability to kill that connection.  It's not exactly an economical attack.  (For contrast, the AES-128-GCM number is ~16 quadrillion.)

> Am I correct in understanding that the practical impact of 
> this boils down to "We might need to do key updates more frequently 
> than we had hoped" since the integrity limits are a bit more modest?

Unfortunately, integrity limits are not reset by a key update, so the net effect is that you have to make a new connection.  But see above.

The confidentiality limits are marginally higher as a result of the new analysis, so you end up having to update a tiny bit less.


From nobody Thu Jun 25 23:50:13 2020
Return-Path: <lars@eggert.org>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 23B193A116D for <quic@ietfa.amsl.com>; Thu, 25 Jun 2020 23:50:12 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.099
X-Spam-Level: 
X-Spam-Status: No, score=-2.099 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=eggert.org
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id yK9DBPwG_t8y for <quic@ietfa.amsl.com>; Thu, 25 Jun 2020 23:50:11 -0700 (PDT)
Received: from mail.eggert.org (mail.eggert.org [IPv6:2a00:ac00:0:35:211:32ff:fe22:186f]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id D1DE33A1161 for <quic@ietf.org>; Thu, 25 Jun 2020 23:50:10 -0700 (PDT)
Received: from [IPv6:2a00:ac00:0:35:2c56:4e30:d5ea:77a4] (unknown [IPv6:2a00:ac00:0:35:2c56:4e30:d5ea:77a4]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.eggert.org (Postfix) with ESMTPSA id E5A49616E08 for <quic@ietf.org>; Fri, 26 Jun 2020 09:13:44 +0300 (EEST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=eggert.org; s=dkim; t=1593152024; bh=uOkMNX44FcDh/Fqforck5dWk//380g4SylmnW2tibpk=; h=From:Subject:Date:To; b=x86AaHVtVHBlEFs9IoO7kkAVjRIbgwBeg6oNgfpI+CytEOeYvLnr6JtVmCJi9fJDO idEDxJy44XI7B5+U/dlhwm/2/JhVCxaPIJZs7CJGJ1efo0yVbwIpbJ3mDmasARMlqf ByPForo5b4J7nVIT7v2Ha/tuO7coQ/iQ1OO9ZIg4=
From: Lars Eggert <lars@eggert.org>
Content-Type: multipart/signed; boundary="Apple-Mail=_BCC58BDB-8472-4410-A66E-DE883F77FC1D"; protocol="application/pgp-signature"; micalg=pgp-sha512
Mime-Version: 1.0 (Mac OS X Mail 13.4 \(3608.80.23.2.2\))
Subject: Doodle poll for -29 virtual interop event
Message-Id: <1A5A359E-7F46-49E0-BCFB-65EE718BBF63@eggert.org>
Date: Fri, 26 Jun 2020 09:13:35 +0300
To: IETF QUIC WG <quic@ietf.org>
X-MailScanner-ID: E5A49616E08.A1E0C
X-MailScanner: Found to be clean
X-MailScanner-From: lars@eggert.org
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/c8DY_kt6FmY8NGRladKmlZHM0QI>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 26 Jun 2020 06:50:12 -0000

--Apple-Mail=_BCC58BDB-8472-4410-A66E-DE883F77FC1D
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=us-ascii

Hi,

please fill out https://doodle.com/poll/rif4gd4ib9he2kqc to find the =
best day(s) for a virtual interop on the -29 drafts.

Thanks,
Lars

--Apple-Mail=_BCC58BDB-8472-4410-A66E-DE883F77FC1D
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
	filename=signature.asc
Content-Type: application/pgp-signature;
	name=signature.asc
Content-Description: Message signed with OpenPGP

-----BEGIN PGP SIGNATURE-----
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=epsO
-----END PGP SIGNATURE-----

--Apple-Mail=_BCC58BDB-8472-4410-A66E-DE883F77FC1D--


From nobody Sun Jun 28 00:32:37 2020
Return-Path: <do_not_reply@mnot.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 17ECC3A0BD7 for <quic@ietfa.amsl.com>; Sun, 28 Jun 2020 00:32:36 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, HTML_MESSAGE=0.001, RCVD_IN_MSPIKE_H3=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=mnot.net header.b=Odo87gFF; dkim=pass (2048-bit key) header.d=messagingengine.com header.b=uVU7MN9Y
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id HyM-Io95BlQ5 for <quic@ietfa.amsl.com>; Sun, 28 Jun 2020 00:32:34 -0700 (PDT)
Received: from out2-smtp.messagingengine.com (out2-smtp.messagingengine.com [66.111.4.26]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id E4D9E3A0BEC for <quic@ietf.org>; Sun, 28 Jun 2020 00:32:33 -0700 (PDT)
Received: from compute1.internal (compute1.nyi.internal [10.202.2.41]) by mailout.nyi.internal (Postfix) with ESMTP id 039195C0090 for <quic@ietf.org>; Sun, 28 Jun 2020 03:32:33 -0400 (EDT)
Received: from mailfrontend2 ([10.202.2.163]) by compute1.internal (MEProxy); Sun, 28 Jun 2020 03:32:33 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=mnot.net; h= content-type:mime-version:from:to:subject:message-id:date; s= fm3; bh=5pHaCQxvGhojXeM/IgXYAwz7RhWTNzPlqxRHzikG/H4=; b=Odo87gFF jkfBXcdLkuCGf1EJvmO1lcQdTAwus3P8ga4zoFU84RlW/WU8YTWecxOLj5Opfq7f mXixKytT49vSGjWisPu21CpUYNFoHZPsvW9CKt04kLVTBAU4OPU6ofZ7a4s8RVKh 1sUYfehR8XRd1iZRm/iahidYwUSA3lXYNhmR5dLdGLh0DhCOnOZhPkcypaXQVNWI CiI7T8ATlpTFxzUPHoEa2ad0esmZVJuavQteecyL5m+R8ZmETDLgGLxQv/dyQ5nQ y8i1IVy0qTBPsDC9OVWBvK/FLkomJCu+wWxgvkjfZpKkjGlSU12yMnkxV+dRxJTw bQPmISi+cRdPVQ==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d= messagingengine.com; h=content-type:date:from:message-id :mime-version:subject:to:x-me-proxy:x-me-proxy:x-me-sender :x-me-sender:x-sasl-enc; s=fm3; bh=5pHaCQxvGhojXeM/IgXYAwz7RhWTN zPlqxRHzikG/H4=; b=uVU7MN9YzvxtGIK2qvBFBQVD7HaOGeUt7M8u/9HpqgAna 7WuLRzleueKCzny8EANxtx4clEX6M0dlZNaZc5fzZWXurwFl+EDeJHxYJFqVmEVH 6VhhRfLWQp9ZtDIz89L8fguBQSQ6tckYZ4meWuGbjXQwh2OTjyzD5pqxsy7PQJRq 5OFiTpE5H1/0NxFYDuVHiJKH2z0QkvUFfAPgus7djFsn5MvjllUshzRZYHlB/DMh sCPgbabQ35/7OYqYdz2b7mBqwhr78STuRurwSkVep9U4DfE/VbN8AicLlSU7tygS rraoOx/llQ5nbNIltgsqRvuHpNqVqHVtKdI26Ed0A==
X-ME-Sender: <xms:kEf4XvPh2KorFLbTJxTJpWUkcPihClUlZmCBEcjpMFXsWbq-IHO61w>
X-ME-Proxy-Cause: gggruggvucftvghtrhhoucdtuddrgeduhedrudelhedgvdejucetufdoteggodetrfdotf fvucfrrhhofhhilhgvmecuhfgrshhtofgrihhlpdfqfgfvpdfurfetoffkrfgpnffqhgen uceurghilhhouhhtmecufedttdenucenucfjughrpegtggfhvffusegrtddtredttdejne cuhfhrohhmpeftvghpohhsihhtohhrhicutegtthhivhhithihucfuuhhmmhgrrhihuceu ohhtuceoughopghnohhtpghrvghplhihsehmnhhothdrnhgvtheqnecuggftrfgrthhtvg hrnhepkeefvdduteejvdefkeehieevuefgfefhteetveegffekffefteffvdelheduieet necuffhomhgrihhnpehgihhthhhusgdrtghomhenucfkphephedvrddvheegrdefrddvvd eknecuvehluhhsthgvrhfuihiivgepudenucfrrghrrghmpehmrghilhhfrhhomhepugho pghnohhtpghrvghplhihsehmnhhothdrnhgvth
X-ME-Proxy: <xmx:kEf4Xp-iUufKZaNk6Lrj2KmVXNsjrGb-O85g0GnFPcSQy4zQP6iYZA> <xmx:kEf4XuRvPoA6dSaWi9rHGyL9r3PXr-UswANKgyU8c4arpQ_YbEU-sg> <xmx:kEf4XjvoL9bfiTCe9uAHkqzEKkkte-6S2bpVHKU2MWvJghhGnPEvZg> <xmx:kEf4XnklzD_BinHUbiLgzZynAKr_rGO5jwK7dPIbKdHI3BR9Vvc3Og>
Received: from fv-az59.internal.cloudapp.net (unknown [52.254.3.228]) by mail.messagingengine.com (Postfix) with ESMTPA id AA5423067AE1 for <quic@ietf.org>; Sun, 28 Jun 2020 03:32:32 -0400 (EDT)
Content-Type: multipart/alternative; boundary="===============5843208928432529992=="
MIME-Version: 1.0
From: Repository Activity Summary Bot <do_not_reply@mnot.net>
To: quic@ietf.org
Subject: Weekly github digest (QUIC Activity Summary)
Message-Id: <20200628073232.AA5423067AE1@mailuser.nyi.internal>
Date: Sun, 28 Jun 2020 03:32:32 -0400 (EDT)
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/WkjPlnQGuXfe4LZGgB7xyQ555ak>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Sun, 28 Jun 2020 07:32:36 -0000

--===============5843208928432529992==
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain; charset="utf-8"; format="flowed"




Events without label "editorial"

Issues
------
* quicwg/base-drafts (+4/-1/=F0=9F=92=AC10)
  4 issues created:
  - New confidentiality and integrity limits (by chris-wood)
    https://github.com/quicwg/base-drafts/issues/3788=20
  - Improve Large Server Certificate Scenario (by nibanks)
    https://github.com/quicwg/base-drafts/issues/3784=20
  - Clarify RESET_STREAM interaction with flow-control limits (by jlaine)
    https://github.com/quicwg/base-drafts/issues/3782=20
  - Minor inconsistency of error when MAX_STREAMS exceeds 2^60 (by LPardue)
    https://github.com/quicwg/base-drafts/issues/3781=20

  6 issues received 10 new comments:
  - #3784 Improve Large Server Certificate Scenario (4 by janaiyengar, mart=
inthomson, nibanks)
    https://github.com/quicwg/base-drafts/issues/3784 [-transport]=20
  - #3782 Clarify RESET_STREAM interaction with flow-control limits (1 by m=
artinthomson)
    https://github.com/quicwg/base-drafts/issues/3782 [-transport]=20
  - #3778 Third ECN validation check and lost acknowledgments (1 by martint=
homson)
    https://github.com/quicwg/base-drafts/issues/3778 [-transport]=20
  - #3765 disable_active_migration with SPA (1 by MikeBishop)
    https://github.com/quicwg/base-drafts/issues/3765 [-transport]=20
  - #3722 QUIC Transport: Flow Control (2 by martinthomson)
    https://github.com/quicwg/base-drafts/issues/3722 [-transport]=20
  - #3690 How to reject a connection attempt (1 by janaiyengar)
    https://github.com/quicwg/base-drafts/issues/3690 [-transport]=20

  1 issues closed:
  - Third ECN validation check and lost acknowledgments https://github.com/=
quicwg/base-drafts/issues/3778 [-transport]=20

* quicwg/datagram (+1/-0/=F0=9F=92=AC14)
  1 issues created:
  - question about scope: why only unreliable datagrams? Why not reliable d=
atagrams too? (by skissane)
    https://github.com/quicwg/datagram/issues/12=20

  2 issues received 14 new comments:
  - #12 Question about scope: why only unreliable datagrams? Why not reliab=
le datagrams too? (4 by DavidSchinazi, Ralith, skissane)
    https://github.com/quicwg/datagram/issues/12=20
  - #6 No streams in datagrams? (10 by LPardue, Ralith, martinduke, nibanks=
, skissane, tfpauly)
    https://github.com/quicwg/datagram/issues/6=20



Pull requests
-------------
* quicwg/base-drafts (+1/-0/=F0=9F=92=AC0)
  1 pull requests submitted:
  - Expand AEAD limits to consider multi-user security. (by chris-wood)
    https://github.com/quicwg/base-drafts/pull/3789=20


Repositories tracked by this digest:
-----------------------------------
* https://github.com/quicwg/base-drafts
* https://github.com/quicwg/ops-drafts
* https://github.com/quicwg/datagram
* https://github.com/quicwg/load-balancers
* https://github.com/quicwg/version-negotiation

--===============5843208928432529992==
Content-Type: text/html; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable

<!doctype html>
<html lang=3D"en">
<head>
<meta charset=3D"utf-8">
<title>Weekly github digest (QUIC Activity Summary)</title>
<style>
body { font-family: Gotham, "Helvetica Neue", Helvetica, Arial, sans-serif;=
 font-size: 14px; }
h2 { margin-top: 3em; color: #A52A2A; font-style: italic; font-weight: norm=
al; }
h3 { margin-bottom:0; margin-top: 2em; font-size: 1.2em; }
h1+h2 { margin-top: 1em; }
a { color: #bb6219; text-decoration: none; }
li { margin-bottom: .35em; }
.repos { margin-bottom: 0; margin-top:0; line-height: 1.2; }
.new { color: red; }
.label { display: inline;
	padding: .2em .6em .3em;
	font-size: 75%;
	font-weight: 700;
	line-height: 1;
	color: #fff;
	text-align: center;
	white-space: nowrap;
	vertical-align: baseline;
	border-radius: .25em;
}
</style>
</head>

<body>
<h1>Sunday June 28, 2020</h1>

<p>Events without label "editorial"</p>

<h2>Issues</h2>

<h3>quicwg/base-drafts (+4/-1/=F0=9F=92=AC10)</h3>
  <p class=3D"new">4 issues created:</p>
  <ul>
  <li>#3788 <a href=3D"https://github.com/quicwg/base-drafts/issues/3788">N=
ew confidentiality and integrity limits</a> (by chris-wood) </li>
 =20
  <li>#3784 <a href=3D"https://github.com/quicwg/base-drafts/issues/3784">I=
mprove Large Server Certificate Scenario</a> (by nibanks) </li>
 =20
  <li>#3782 <a href=3D"https://github.com/quicwg/base-drafts/issues/3782">C=
larify RESET_STREAM interaction with flow-control limits</a> (by jlaine) </=
li>
 =20
  <li>#3781 <a href=3D"https://github.com/quicwg/base-drafts/issues/3781">M=
inor inconsistency of error when MAX_STREAMS exceeds 2^60</a> (by LPardue) =
</li>
  </ul>

  <p>6 issues received 10 new comments:</p>
  <ul>
  <li>#3784 <a href=3D"https://github.com/quicwg/base-drafts/issues/3784">I=
mprove Large Server Certificate Scenario</a> (4 by janaiyengar, martinthoms=
on, nibanks) <span class=3D"label" style=3D"background-color: #fbca04; colo=
r: #000000">-transport</span> </li>
 =20
  <li>#3782 <a href=3D"https://github.com/quicwg/base-drafts/issues/3782">C=
larify RESET_STREAM interaction with flow-control limits</a> (1 by martinth=
omson) <span class=3D"label" style=3D"background-color: #fbca04; color: #00=
0000">-transport</span> </li>
 =20
  <li>#3778 <a href=3D"https://github.com/quicwg/base-drafts/issues/3778">T=
hird ECN validation check and lost acknowledgments</a> (1 by martinthomson)=
 <span class=3D"label" style=3D"background-color: #fbca04; color: #000000">=
-transport</span> </li>
 =20
  <li>#3765 <a href=3D"https://github.com/quicwg/base-drafts/issues/3765">d=
isable_active_migration with SPA</a> (1 by MikeBishop) <span class=3D"label=
" style=3D"background-color: #fbca04; color: #000000">-transport</span> </l=
i>
 =20
  <li>#3722 <a href=3D"https://github.com/quicwg/base-drafts/issues/3722">Q=
UIC Transport: Flow Control</a> (2 by martinthomson) <span class=3D"label" =
style=3D"background-color: #fbca04; color: #000000">-transport</span> </li>
 =20
  <li>#3690 <a href=3D"https://github.com/quicwg/base-drafts/issues/3690">H=
ow to reject a connection attempt</a> (1 by janaiyengar) <span class=3D"lab=
el" style=3D"background-color: #fbca04; color: #000000">-transport</span> <=
/li>
  </ul>

  <p>1 issues closed:</p>
  <ul>
  <li>#3778 <a href=3D"https://github.com/quicwg/base-drafts/issues/3778">T=
hird ECN validation check and lost acknowledgments</a> <span class=3D"label=
" style=3D"background-color: #fbca04; color: #000000">-transport</span> </l=
i>
  </ul>

<h3>quicwg/datagram (+1/-0/=F0=9F=92=AC14)</h3>
  <p class=3D"new">1 issues created:</p>
  <ul>
  <li>#12 <a href=3D"https://github.com/quicwg/datagram/issues/12">question=
 about scope: why only unreliable datagrams? Why not reliable datagrams too=
?</a> (by skissane) </li>
  </ul>

  <p>2 issues received 14 new comments:</p>
  <ul>
  <li>#12 <a href=3D"https://github.com/quicwg/datagram/issues/12">Question=
 about scope: why only unreliable datagrams? Why not reliable datagrams too=
?</a> (4 by DavidSchinazi, Ralith, skissane) </li>
 =20
  <li>#6 <a href=3D"https://github.com/quicwg/datagram/issues/6">No streams=
 in datagrams?</a> (10 by LPardue, Ralith, martinduke, nibanks, skissane, t=
fpauly) </li>
  </ul>




<h2>Pull requests</h2>
<h3>quicwg/base-drafts (+1/-0/=F0=9F=92=AC0)</h3>
  <p class=3D"new">1 pull requests submitted:</p>
  <ul>
  <li>#3789 <a href=3D"https://github.com/quicwg/base-drafts/pull/3789">Exp=
and AEAD limits to consider multi-user security.</a> (by chris-wood) </li>
  </ul>




<h2>Repositories tracked by this digest:</h2>
<ul class=3D"repos">
  <li><a href=3D"https://github.com/quicwg/base-drafts">https://github.com/=
quicwg/base-drafts</a></li>
  <li><a href=3D"https://github.com/quicwg/ops-drafts">https://github.com/q=
uicwg/ops-drafts</a></li>
  <li><a href=3D"https://github.com/quicwg/datagram">https://github.com/qui=
cwg/datagram</a></li>
  <li><a href=3D"https://github.com/quicwg/load-balancers">https://github.c=
om/quicwg/load-balancers</a></li>
  <li><a href=3D"https://github.com/quicwg/version-negotiation">https://git=
hub.com/quicwg/version-negotiation</a></li>
  </ul>
</body>
</html>

--===============5843208928432529992==--


From nobody Mon Jun 29 23:55:21 2020
Return-Path: <lars@eggert.org>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id D45893A0F3D for <quic@ietfa.amsl.com>; Mon, 29 Jun 2020 23:55:19 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.099
X-Spam-Level: 
X-Spam-Status: No, score=-2.099 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=eggert.org
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Lv_HC4ITCPjk for <quic@ietfa.amsl.com>; Mon, 29 Jun 2020 23:55:18 -0700 (PDT)
Received: from mail.eggert.org (mail.eggert.org [91.190.195.94]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 3CFF83A0F3C for <quic@ietf.org>; Mon, 29 Jun 2020 23:55:18 -0700 (PDT)
Received: from [IPv6:2a00:ac00:0:35:c482:f129:80b6:7a52] (unknown [IPv6:2a00:ac00:0:35:c482:f129:80b6:7a52]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.eggert.org (Postfix) with ESMTPSA id 15A6661169D for <quic@ietf.org>; Tue, 30 Jun 2020 09:55:07 +0300 (EEST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=eggert.org; s=dkim; t=1593500108; bh=jyvdokztEBiDFq7ihSWDYkAdbh8drvvlmvOgKcPhB5I=; h=From:Reply-To:Subject:Date:To; b=GyPwZS5Q21x9NarV+BN/5LgYn7hn/n3U2wYsDJddd3jM/R9WN9tJ1o17rpr+/MA/1 cuW7ATGLtzx6+Tt/fZh5x8jdKH6OVhqQ2A2j2RyfebqsND9DkzsYJGHzHqz+NaxkdW L1yHPLXDn7hp89hknbRaAlrr38MqP6dFhvUZXQpo=
From: Lars Eggert <lars@eggert.org>
Content-Type: multipart/signed; boundary="Apple-Mail=_6FD7F95F-3E26-4F2F-BE2D-8820F9749139"; protocol="application/pgp-signature"; micalg=pgp-sha512
Reply-To: WG Chairs <quic-chairs@ietf.org>
Mime-Version: 1.0 (Mac OS X Mail 13.4 \(3608.80.23.2.2\))
Subject: Please send agenda request for IETF-108
Message-Id: <81DED556-AAAB-4833-A0CE-5F4DA2AE4869@eggert.org>
Date: Tue, 30 Jun 2020 09:55:07 +0300
To: IETF QUIC WG <quic@ietf.org>
X-MailScanner-ID: 15A6661169D.A0940
X-MailScanner: Found to be clean
X-MailScanner-From: lars@eggert.org
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/pQgSUbVLcyF8H8vJoH8BsZh2HnU>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 30 Jun 2020 06:55:20 -0000

--Apple-Mail=_6FD7F95F-3E26-4F2F-BE2D-8820F9749139
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=us-ascii

Hi,

please send your agenda requests for IETF-108 to quic-chairs@ietf.org.

Please note that we'll strictly prioritize items that will let us =
progress the drafts currently in WGLC towards IETF LC, followed by =
requests related to adopted WG drafts, followed by everything else. As a =
tie breaker, we'll prioritize things that have seen active discussion in =
the recent past.

Thanks,
Lars

--Apple-Mail=_6FD7F95F-3E26-4F2F-BE2D-8820F9749139
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
	filename=signature.asc
Content-Type: application/pgp-signature;
	name=signature.asc
Content-Description: Message signed with OpenPGP

-----BEGIN PGP SIGNATURE-----
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=nP5m
-----END PGP SIGNATURE-----

--Apple-Mail=_6FD7F95F-3E26-4F2F-BE2D-8820F9749139--


From tim.jebb@bt.com  Tue Jun 30 03:16:07 2020
Return-Path: <tim.jebb@bt.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 9962C3A1173 for <quic@ietfa.amsl.com>; Tue, 30 Jun 2020 03:16:07 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.097
X-Spam-Level: 
X-Spam-Status: No, score=-2.097 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, HTML_MESSAGE=0.001, RCVD_IN_MSPIKE_H4=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=bt.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id CuO5IqYsTeOw for <quic@ietfa.amsl.com>; Tue, 30 Jun 2020 03:16:02 -0700 (PDT)
Received: from smtpe1.intersmtp.com (smtpe1.intersmtp.com [213.121.35.71]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 42E133A116A for <quic@ietf.org>; Tue, 30 Jun 2020 03:16:00 -0700 (PDT)
Received: from tpw09926dag11e.domain1.systemhost.net (10.9.212.11) by BWP09926076.bt.com (10.36.82.107) with Microsoft SMTP Server (version=TLS1_2,  cipher=TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA384_P256) id 15.1.1713.5; Tue, 30 Jun 2020 11:15:33 +0100
Received: from tpw09926dag08h.domain1.systemhost.net (10.9.202.47) by tpw09926dag11e.domain1.systemhost.net (10.9.212.11) with Microsoft SMTP Server (TLS) id 15.0.1395.4; Tue, 30 Jun 2020 11:15:58 +0100
Received: from bwp09926080.bt.com (10.36.82.111) by tpw09926dag08h.domain1.systemhost.net (10.9.202.47) with Microsoft SMTP Server (TLS) id 15.0.1395.4 via Frontend Transport; Tue, 30 Jun 2020 11:15:57 +0100
Received: from GBR01-CWL-obe.outbound.protection.outlook.com (104.47.20.56) by smtpe1.intersmtp.com (10.36.82.111) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA384_P256) id 15.1.1713.5; Tue, 30 Jun 2020 11:15:37 +0100
ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=iy1rQP8MDR1kIDBHPDTiYYRzVq5OwTgc96YGEjLuRRrk92nSn56XVc2D/LBbUfZ2+iIs5+CegCmwnqCBcnco9DAIdJKVSpqZ+b+6JNFSK4d1iJPLsV0W2a7t/PQxLBDrtaz/SsSc3RAi4vF3vAPo3FMb55cKQx2bEaqo0Jd96fU7NHoDZZlEhhiGSBlIE1Y4u0pFi8OxiNh1MFAlx6qLaNkdsq9FULD+BoDpKL85qJMtNR9CtTgoGCbtzLQ9gxZHcmsqbkJyb6CrjKqgT4zG5Wovr7rtpW5HqFG2wwsjVmcjwoRsKj25w4aFgTOCqU+9ADts0tbYcegIj2elpkjCKA==
ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com;  s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=DlJGdPD9XG1LzA25SR1Zdt9GK0XTQEypdSz3xRmdPUU=; b=LXilIzuTnjnfIrj92nPiaQjuFU+4cNiweEQOpMxJIodDyMKfb25tW/ZFVjlqBTBAFWKGijWxTzEE0G25OPwvnSSBSZWEJ3nxlc+AiWTzkNv5k0RnHquaqP99eCzIILqn8tkrXF5coVuKuJltDa2cQJlkzFxUXNdDBWU3/9znYiztzQ+6VXGlTampfaWcd8oF7Op+KGFwr9u6K4CwSN5/tXlxfUhW79PsgNiSunv6m/6Ec0by9mcnfST2Kp3oXIBfLxP9kLLPItuie91gLc8/W9muLWxtwwA0Kym79wkaOcIun5Bo+ilzEjujLCykWPaFIBtt8t6iHCxirNuykHY7+w==
ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=bt.com; dmarc=pass action=none header.from=bt.com; dkim=pass header.d=bt.com; arc=none
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=bt.com; s=selector1; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=DlJGdPD9XG1LzA25SR1Zdt9GK0XTQEypdSz3xRmdPUU=; b=W/jk3t0+eEv5A36W/kHxy51MegzwKpwYT+EVQtUb+gypvCr/SZfAT/lsyBMtPz4sSZ4Aq4N0m+ye1nmgDj/9I1pYu2i1axuKiHuatODI/JK9cBYP1vOkl/EbZlrYvpqJFDKV+GMB+dSiQ01bTx2cDlzlm4mSXOvSyn9EprqjWEA=
Received: from LNXP123MB2460.GBRP123.PROD.OUTLOOK.COM (2603:10a6:600:dc::10) by LO2P123MB2096.GBRP123.PROD.OUTLOOK.COM (2603:10a6:600:c4::13) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.3153.20; Tue, 30 Jun 2020 10:15:53 +0000
Received: from LNXP123MB2460.GBRP123.PROD.OUTLOOK.COM ([fe80::ec07:dcb5:1747:4632]) by LNXP123MB2460.GBRP123.PROD.OUTLOOK.COM ([fe80::ec07:dcb5:1747:4632%7]) with mapi id 15.20.3131.028; Tue, 30 Jun 2020 10:15:53 +0000
From: <tim.jebb@bt.com>
To: <quic@ietf.org>
Subject: Padding in Initial Packets
Thread-Topic: Padding in Initial Packets
Thread-Index: AdZOuPMsC8HOn4IhREy6+3yKJlq+/g==
Date: Tue, 30 Jun 2020 10:15:52 +0000
Message-ID: <LNXP123MB24602880B0AE644B9B1F3F83F26F0@LNXP123MB2460.GBRP123.PROD.OUTLOOK.COM>
Accept-Language: en-GB, en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
authentication-results: ietf.org; dkim=none (message not signed) header.d=none;ietf.org; dmarc=none action=none header.from=bt.com;
x-originating-ip: [146.199.174.244]
x-ms-publictraffictype: Email
x-ms-office365-filtering-correlation-id: b54a9350-303f-4871-808e-08d81cde9209
x-ms-traffictypediagnostic: LO2P123MB2096:
x-microsoft-antispam-prvs: <LO2P123MB2096F8824A68FF0A81D41609F26F0@LO2P123MB2096.GBRP123.PROD.OUTLOOK.COM>
x-antispam-2: 1
x-ms-oob-tlc-oobclassifiers: OLM:8273;
x-forefront-prvs: 0450A714CB
x-ms-exchange-senderadcheck: 1
x-microsoft-antispam: BCL:0;
x-microsoft-antispam-message-info: NUHN7Xj4EdMbeF5QBMVoi4POlKOzDw6rOijxkcDgx9c+o0lRocaqLwokYMO0kDoSmUK/NHuRC6qYXRF7z9PFNrQFO0lTWJb/BZn9xZStQmfS56RlfC4Wcz/Nrg7e3rHHZxbUbNk4CpRzvBwU74E9Qf5/vYSwpEzJWoi/fyNNH4JyEuvZ7tLzPbALRk6olaX5anQmBTPjsIaVf5dqoOdYmxifH4FGTgdPbhEoRjLcuk9OzP3UVWIjAUXyaIUjpOayer7/BQsFsp9d1Vbv/B8eIgt5zqyETzBz5IFcN3cqFdoNeHYGAmjIowF9A+wrMdDUSZkEl/BcXr9zRqC9+z6oqg==
x-forefront-antispam-report: CIP:255.255.255.255; CTRY:; LANG:en; SCL:1; SRV:;  IPV:NLI; SFV:NSPM; H:LNXP123MB2460.GBRP123.PROD.OUTLOOK.COM; PTR:; CAT:NONE;  SFTY:; SFS:(4636009)(346002)(396003)(39860400002)(376002)(366004)(136003)(26005)(2906002)(5660300002)(6506007)(9326002)(8936002)(66476007)(64756008)(66556008)(66446008)(66946007)(8676002)(71200400001)(86362001)(9686003)(3480700007)(186003)(83380400001)(55016002)(7696005)(76116006)(478600001)(52536014)(33656002)(6916009)(316002); DIR:OUT; SFP:1101; 
x-ms-exchange-antispam-messagedata: NqRkKzMLJ2J/ZIsw2nf/MRntOTVSofYGqorHo70NDTokBK3LRHVn5DWhwAblnTgwjbr/gFVTZJJNXlx8mZXGk6xwHSWBg8RD+OzGi3IeRCXu5CnMizgkNLi1p7niklduKZj9kzjLIYXUAMzQoUeYVu81v7PxrjJpMlLMhrpoiP9xto5aeT57yIin3ZsKLnV6DAMmLFmS4LqXf9naszKNiVqwNdX7031obB9r3rdf8u0/aTLTQK6FSP3+iSn+BTkEBY2isewMthPxF/oVHrcgnAbjqjcAzCu9i3SBdtiZq7n+PC8TPQc49x3U0fhKx3hKgf9fYS+7yZv8Bsaa/PDxi7k8pyZWw0vlAqYXUUre63NxYDFoM9qj+hCv1XKm5TMPHXUuq2brla+qrk7YHfw0iRCwmkSmFP6KeMkc0e/U6ZDa6WUdScLQk7b8O+SaInMudzmtMvXo3XKMudOyjYA7hHbk43KCN0DyeXlzZxcpJWWldKiQ8a6611w0NcwerX8A
x-ms-exchange-transport-forked: True
Content-Type: multipart/alternative; boundary="_000_LNXP123MB24602880B0AE644B9B1F3F83F26F0LNXP123MB2460GBRP_"
MIME-Version: 1.0
X-MS-Exchange-CrossTenant-AuthAs: Internal
X-MS-Exchange-CrossTenant-AuthSource: LNXP123MB2460.GBRP123.PROD.OUTLOOK.COM
X-MS-Exchange-CrossTenant-Network-Message-Id: b54a9350-303f-4871-808e-08d81cde9209
X-MS-Exchange-CrossTenant-originalarrivaltime: 30 Jun 2020 10:15:53.0359 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Hosted
X-MS-Exchange-CrossTenant-id: a7f35688-9c00-4d5e-ba41-29f146377ab0
X-MS-Exchange-CrossTenant-mailboxtype: HOSTED
X-MS-Exchange-CrossTenant-userprincipalname: 5+5t5Yt2lhaplWn3djpgW48K78RtZWKEMCC5smli+9xztiwZv3wolqaWsE8dXfku
X-MS-Exchange-Transport-CrossTenantHeadersStamped: LO2P123MB2096
X-OriginatorOrg: bt.com
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/7y9-Y9vw0pIYdDDtKYl9caXBFIk>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 30 Jun 2020 10:18:10 -0000

--_000_LNXP123MB24602880B0AE644B9B1F3F83F26F0LNXP123MB2460GBRP_
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

Hi All,

I'm fairly new to QUIC so apologies if the questions in here are dumb.

I've been going through a QUIC pcap, which I captured from a Firefox implem=
entation of IETF draft 27 QUIC, and am a little confused about the use of p=
adding in the first Initial packet.  According to Wireshark,  the UDP datag=
ram contains a QUIC packet of type initial, length 272, the payload of whic=
h is a Client Hello.

The remainder of the UDP datagram is 1055 zero bytes.  According to Wiresha=
rk these zero bytes constitute a QUIC IETF Packet.

However there doesn't seem to be a valid QUIC packet header for these bytes=
.  AIUI according to the QUIC spec padding bytes are frames within a packet=
, so these zero bytes seem to be just bytes in the UDP datagram outside of =
the QUIC payload.  Which would be fine as far as I can see, but:

Why does Wireshark say it's QUIC IETF?
Is there anything in any of the QUIC draft RFCs mandating this behaviour, i=
f so where?
Why is it being done - can't see the point?
If padding is to be added surely it should be as part of a QUIC packet and =
encrypted - here there's just a load of plaintext zero bytes.


Here is Wireshark's analysis:

"QUIC IETF
    QUIC Connection information
    [Packet Length: 302]
    1... .... =3D Header Form: Long Header (1)
    .1.. .... =3D Fixed Bit: True
    ..00 .... =3D Packet Type: Initial (0)
    .... 00.. =3D Reserved: 0
    .... ..10 =3D Packet Number Length: 3 bytes (2)
    Version: draft-27 (0xff00001b)
    Destination Connection ID Length: 17
    Destination Connection ID: 76f81a5588b2f25c5ab6dbd461a8d749e1
    Source Connection ID Length: 3
    Source Connection ID: c5ef02
    Token Length: 0
    Length: 272
    Packet Number: 0
    Payload: 04f535177fcbcfd5f06d24ffde66edf10df79205666b1851...
    TLSv1.3 Record Layer: Handshake Protocol: Client Hello
        Frame Type: CRYPTO (0x0000000000000006)
        Offset: 0
        Length: 249
        Crypto Data
        Handshake Protocol: Client Hello
            Handshake Type: Client Hello (1)
            Length: 245
            Version: TLS 1.2 (0x0303)
            Random: b1b76126ddab3c639cdfaa7d92d6cabe2ea9c1a19ea86b8b...
            Session ID Length: 0
            Cipher Suites Length: 4
            Cipher Suites (2 suites)
            Compression Methods Length: 1
            Compression Methods (1 method)
            Extensions Length: 200
            Extension: server_name (len=3D15)
            Extension: extended_master_secret (len=3D0)
            Extension: renegotiation_info (len=3D1)
            Extension: supported_groups (len=3D20)
            Extension: application_layer_protocol_negotiation (len=3D8)
            Extension: status_request (len=3D5)
            Extension: key_share (len=3D38)
            Extension: supported_versions (len=3D3)
            Extension: signature_algorithms (len=3D16)
            Extension: psk_key_exchange_modes (len=3D2)
            Extension: record_size_limit (len=3D2)
            Extension: quic_transports_parameters (len=3D42)
QUIC IETF
    [Packet Length: 1055]
    Remaining Payload: 000000000000000000000000000000000000000000000000..."

Any comments or explanations?

Tim

--_000_LNXP123MB24602880B0AE644B9B1F3F83F26F0LNXP123MB2460GBRP_
Content-Type: text/html; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

<html xmlns:v=3D"urn:schemas-microsoft-com:vml" xmlns:o=3D"urn:schemas-micr=
osoft-com:office:office" xmlns:w=3D"urn:schemas-microsoft-com:office:word" =
xmlns:m=3D"http://schemas.microsoft.com/office/2004/12/omml" xmlns=3D"http:=
//www.w3.org/TR/REC-html40">
<head>
<meta http-equiv=3D"Content-Type" content=3D"text/html; charset=3Dus-ascii"=
>
<meta name=3D"Generator" content=3D"Microsoft Word 15 (filtered medium)">
<style><!--
/* Font Definitions */
@font-face
	{font-family:"Cambria Math";
	panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
	{font-family:Calibri;
	panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
	{margin:0cm;
	margin-bottom:.0001pt;
	font-size:11.0pt;
	font-family:"Calibri",sans-serif;
	mso-fareast-language:EN-US;}
a:link, span.MsoHyperlink
	{mso-style-priority:99;
	color:#0563C1;
	text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
	{mso-style-priority:99;
	color:#954F72;
	text-decoration:underline;}
span.EmailStyle17
	{mso-style-type:personal-compose;
	font-family:"Calibri",sans-serif;
	color:windowtext;}
.MsoChpDefault
	{mso-style-type:export-only;
	font-family:"Calibri",sans-serif;
	mso-fareast-language:EN-US;}
@page WordSection1
	{size:612.0pt 792.0pt;
	margin:72.0pt 72.0pt 72.0pt 72.0pt;}
div.WordSection1
	{page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext=3D"edit" spidmax=3D"1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext=3D"edit">
<o:idmap v:ext=3D"edit" data=3D"1" />
</o:shapelayout></xml><![endif]-->
</head>
<body lang=3D"EN-GB" link=3D"#0563C1" vlink=3D"#954F72">
<div class=3D"WordSection1">
<p class=3D"MsoNormal">Hi All,<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">I&#8217;m fairly new to QUIC so apologies if the que=
stions in here are dumb.<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">I&#8217;ve been going through a QUIC pcap, which I c=
aptured from a Firefox implementation of IETF draft 27 QUIC, and am a littl=
e confused about the use of padding in the first Initial packet.&nbsp; Acco=
rding to Wireshark,&nbsp; the UDP datagram contains
 a QUIC packet of type initial, length 272, the payload of which is a Clien=
t Hello.&nbsp;
<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">The remainder of the UDP datagram is 1055 zero bytes=
.&nbsp; According to Wireshark these zero bytes constitute a QUIC IETF Pack=
et.<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">However there doesn&#8217;t seem to be a valid QUIC =
packet header for these bytes.&nbsp; AIUI according to the QUIC spec paddin=
g bytes are frames within a packet, so these zero bytes seem to be just byt=
es in the UDP datagram outside of the QUIC payload.&nbsp;
 Which would be fine as far as I can see, but:<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">Why does Wireshark say it&#8217;s QUIC IETF?<o:p></o=
:p></p>
<p class=3D"MsoNormal">Is there anything in any of the QUIC draft RFCs mand=
ating this behaviour, if so where?<o:p></o:p></p>
<p class=3D"MsoNormal">Why is it being done &#8211; can&#8217;t see the poi=
nt?<o:p></o:p></p>
<p class=3D"MsoNormal">If padding is to be added surely it should be as par=
t of a QUIC packet and encrypted &#8211; here there&#8217;s just a load of =
plaintext zero bytes.<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">Here is Wireshark&#8217;s analysis:<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">&#8220;QUIC IETF<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; QUIC Connection information<o:p><=
/o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; [Packet Length: 302]<o:p></o:p></=
p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; 1... .... =3D Header Form: Long H=
eader (1)<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; .1.. .... =3D Fixed Bit: True<o:p=
></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; ..00 .... =3D Packet Type: Initia=
l (0)<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; .... 00.. =3D Reserved: 0<o:p></o=
:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; .... ..10 =3D Packet Number Lengt=
h: 3 bytes (2)<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; Version: draft-27 (0xff00001b)<o:=
p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; Destination Connection ID Length:=
 17<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; Destination Connection ID: 76f81a=
5588b2f25c5ab6dbd461a8d749e1<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; Source Connection ID Length: 3<o:=
p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; Source Connection ID: c5ef02<o:p>=
</o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; Token Length: 0<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; Length: 272<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; Packet Number: 0<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; Payload: 04f535177fcbcfd5f06d24ff=
de66edf10df79205666b1851&#8230;<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; TLSv1.3 Record Layer: Handshake P=
rotocol: Client Hello<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Frame Typ=
e: CRYPTO (0x0000000000000006)<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Offset: 0=
<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Length: 2=
49<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Crypto Da=
ta<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Handshake=
 Protocol: Client Hello<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Handshake Type: Client Hello (1)<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Length: 245<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Version: TLS 1.2 (0x0303)<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Random: b1b76126ddab3c639cdfaa7d92d6cabe2ea9c1a19ea86b8b&#82=
30;<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Session ID Length: 0<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Cipher Suites Length: 4<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Cipher Suites (2 suites)<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Compression Methods Length: 1<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Compression Methods (1 method)<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Extensions Length: 200<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Extension: server_name (len=3D15)<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Extension: extended_master_secret (len=3D0)<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Extension: renegotiation_info (len=3D1)<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Extension: supported_groups (len=3D20)<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Extension: application_layer_protocol_negotiation (len=3D8)<=
o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Extension: status_request (len=3D5)<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Extension: key_share (len=3D38)<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Extension: supported_versions (len=3D3)<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Extension: signature_algorithms (len=3D16)<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Extension: psk_key_exchange_modes (len=3D2)<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Extension: record_size_limit (len=3D2)<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp; Extension: quic_transports_parameters (len=3D42)<o:p></o:p><=
/p>
<p class=3D"MsoNormal">QUIC IETF<o:p></o:p></p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; [Packet Length: 1055]<o:p></o:p><=
/p>
<p class=3D"MsoNormal">&nbsp;&nbsp;&nbsp; Remaining Payload: 00000000000000=
0000000000000000000000000000000000&#8230;&#8221;<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">Any comments or explanations?<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">Tim<o:p></o:p></p>
</div>
</body>
</html>

--_000_LNXP123MB24602880B0AE644B9B1F3F83F26F0LNXP123MB2460GBRP_--


From nobody Tue Jun 30 09:16:54 2020
Return-Path: <mbishop@evequefou.be>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 2D8453A0C07 for <quic@ietfa.amsl.com>; Tue, 30 Jun 2020 09:16:53 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.9
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, HTML_MESSAGE=0.001, RCVD_IN_MSPIKE_H2=-0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=evequefou.onmicrosoft.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 0upyZlUh3WnG for <quic@ietfa.amsl.com>; Tue, 30 Jun 2020 09:16:51 -0700 (PDT)
Received: from NAM11-DM6-obe.outbound.protection.outlook.com (mail-dm6nam11on2105.outbound.protection.outlook.com [40.107.223.105]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 0E68D3A0C01 for <quic@ietf.org>; Tue, 30 Jun 2020 09:16:50 -0700 (PDT)
ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=a5JpoBk5UY90b0AmDsD6HWLZCjw5j3O3/MMjvWUbJIzNLUa63hRBrJ8cwQO2bi+XJQmMQMf9R7ZNAP2yAIR2Pg0gkcIy0BUUBEWw3d0azXqCPkZDRhzl2obM23/tLZur+s3zwrk48AtI1H5NClZeBOr+iyJu6oqRJjVwXUTugQAPc5y1HHaij1AswImqsuvW4JLfWZxZM1/6WR/mZ2KKeS0ZbDBENgeT/Z7Pr7kMKNISMnNbhq8kv6queDRieBxoOBzN8cpAnzsx4Tc0Tc3GwbCBcYrfv/GjhSsqeAm9WeoEKyqF4rgDMcrnfKSIzY9BieFx4RonG1Iy1gtFQUTmVA==
ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com;  s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=Uy66gRlSqCiqYHdQjg7lm/iZb1BgrFgbuscclM+TpIg=; b=iRQsvMPa5ZPgZ6mL08AF4XfSlh0jCmvvxQhD6lJ3KpRSf5yU9/x4138O+vI/PasXpwqStDKFPysWE4YAIKwKSQkB6hBtjVDEXtuTQECZtXgFKF4gEzYdOM1g0uQxiiVQST7vseP1KzOKzyEoIiK2yIv7UBymtn/NqIM+ReHnGDSEnjr+6wxVCBaFm3QIYVHlE2jD4OPSMaCXsdiuKxesB7pyiKjVy3++kqW0U6dvpFqXnOyqKWTjOKgJ1xDO8Ud0zOcmyhdiKXiUX9kk8FleaeFwN13kSz93ILZbZ8w7Bd5AFJw1EzQHXL+WZkp2PzlscpOTYvx6rSqGsZ9NyGYmXQ==
ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=evequefou.be; dmarc=pass action=none header.from=evequefou.be; dkim=pass header.d=evequefou.be; arc=none
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=evequefou.onmicrosoft.com; s=selector2-evequefou-onmicrosoft-com; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=Uy66gRlSqCiqYHdQjg7lm/iZb1BgrFgbuscclM+TpIg=; b=aeNod0XpsfOb/gC5ZUpXdU5iMIugMRzY51eQy22RAwaJ/9q1zjW1GkxK6v8sva0YoDmYq0U14/dKFmvZDIAkqljuxwkraowJ2DHcN03OPF+ylw9sXIMlVGLKiUBBeYGaBXn7N7HONgJTKfbhCz5SAySx/FfwdGNSGK7Om+v+Mrk=
Received: from CH2PR22MB2086.namprd22.prod.outlook.com (2603:10b6:610:8c::8) by CH2PR22MB1847.namprd22.prod.outlook.com (2603:10b6:610:8e::14) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.3153.20; Tue, 30 Jun 2020 16:16:47 +0000
Received: from CH2PR22MB2086.namprd22.prod.outlook.com ([fe80::f0e6:8202:6056:a596]) by CH2PR22MB2086.namprd22.prod.outlook.com ([fe80::f0e6:8202:6056:a596%3]) with mapi id 15.20.3131.027; Tue, 30 Jun 2020 16:16:47 +0000
From: Mike Bishop <mbishop@evequefou.be>
To: "tim.jebb@bt.com" <tim.jebb@bt.com>, "quic@ietf.org" <quic@ietf.org>
Subject: RE: Padding in Initial Packets
Thread-Topic: Padding in Initial Packets
Thread-Index: AdZOuPMsC8HOn4IhREy6+3yKJlq+/gAP8oOg
Date: Tue, 30 Jun 2020 16:16:47 +0000
Message-ID: <CH2PR22MB20869C49226FFDD90CED291DDA6F0@CH2PR22MB2086.namprd22.prod.outlook.com>
References: <LNXP123MB24602880B0AE644B9B1F3F83F26F0@LNXP123MB2460.GBRP123.PROD.OUTLOOK.COM>
In-Reply-To: <LNXP123MB24602880B0AE644B9B1F3F83F26F0@LNXP123MB2460.GBRP123.PROD.OUTLOOK.COM>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
authentication-results: bt.com; dkim=none (message not signed) header.d=none;bt.com; dmarc=none action=none header.from=evequefou.be;
x-originating-ip: [72.49.212.17]
x-ms-publictraffictype: Email
x-ms-office365-filtering-correlation-id: 55c91647-fcb4-40b7-df19-08d81d10fd20
x-ms-traffictypediagnostic: CH2PR22MB1847:
x-microsoft-antispam-prvs: <CH2PR22MB184795E623C3A6EFCBD75C43DA6F0@CH2PR22MB1847.namprd22.prod.outlook.com>
x-ms-oob-tlc-oobclassifiers: OLM:6430;
x-forefront-prvs: 0450A714CB
x-ms-exchange-senderadcheck: 1
x-microsoft-antispam: BCL:0;
x-microsoft-antispam-message-info: RaP6PU0oK7rP2Ia3dSroo/xKWIBeNOHLnSjtX5KGL5CzrknL9QjFfbL1RdEgI7G4i5cucjeHQrG/cnPPDA1HkJjrwJwrcG8ZjjYsFROyMMYW8MtVGVllB66je5in9LR86TqI2eftJWCzNTV1DT/KOd0vxiBVa3tryXLZj33kC48djSqKZ5Tm76wKusvGMjOrip+zjQl2Sr4oqzU7iK5e9PhPG3gaI+L1KSAXc6AGwtZrtaIlRxY6VMPFpEqj6OUaI4GRvwfdbdaiH6lBSbdPV+wsjBVGA1TO/74MJofpLPJpzBol7/HXlHVSCOr0ZsrrBywElEcMAIfjxKzhaHwBHg==
x-forefront-antispam-report: CIP:255.255.255.255; CTRY:; LANG:en; SCL:1; SRV:;  IPV:NLI; SFV:NSPM; H:CH2PR22MB2086.namprd22.prod.outlook.com; PTR:; CAT:NONE;  SFTY:; SFS:(366004)(346002)(66556008)(64756008)(66446008)(76116006)(71200400001)(66476007)(66946007)(83380400001)(508600001)(8676002)(7696005)(110136005)(9686003)(52536014)(55016002)(33656002)(2906002)(26005)(8936002)(186003)(86362001)(5660300002)(3480700007)(53546011)(6506007); DIR:OUT; SFP:1102; 
x-ms-exchange-antispam-messagedata: 6dD16f5hRMbtm+IyBl2ZbNaDlD63OaVcxBO7bA0z+AvR4ZJetIhwDB7kzawr/D1nTeEB7X6Ly3Trb5OSxN9qWsIsJTTinE2/ZxFZLpo6DRslmyorbQX6L1bHuvFzBBkmshklVaatlS9XIODOOHJOOTUppm2eHBfCmgzAdL923dS59AlVYj6Irzu4jMJt6ck3lNYLSqdqFC+JcWA/XAm17rdrskN3PmUQfroO5OS+CPt0IUX1cAGQ9CkQx0vA4WgoF8AgG9rEerZ97jXM3lYYy6r1o19TY6dJh2AbserSj0F9rxz/0VLMAhk5Y/W4ohgmIIp8Mol67RNgAbZ3rOB8cjYzgJseSr0hTllK5wAz0FSPtCEJw2HZyxtFofclvoVmKCMgRVdUKb67hf+wnaDZU2X02BWjNjEQsPP5WgpCeEMPZJlcuSGct3D+2dxAKytGPQc30k2zOGkveRhfl2XrjENi8ZLyLWUS2hEGOE41PXI=
x-ms-exchange-transport-forked: True
Content-Type: multipart/alternative; boundary="_000_CH2PR22MB20869C49226FFDD90CED291DDA6F0CH2PR22MB2086namp_"
MIME-Version: 1.0
X-OriginatorOrg: evequefou.be
X-MS-Exchange-CrossTenant-AuthAs: Internal
X-MS-Exchange-CrossTenant-AuthSource: CH2PR22MB2086.namprd22.prod.outlook.com
X-MS-Exchange-CrossTenant-Network-Message-Id: 55c91647-fcb4-40b7-df19-08d81d10fd20
X-MS-Exchange-CrossTenant-originalarrivaltime: 30 Jun 2020 16:16:47.4784 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Hosted
X-MS-Exchange-CrossTenant-id: 41eaf50b-882d-47eb-8c4c-0b5b76a9da8f
X-MS-Exchange-CrossTenant-mailboxtype: HOSTED
X-MS-Exchange-CrossTenant-userprincipalname: m4O+2C/t2idhmQ73rW7vsSFfS6n7aEKvu59KoEvV+YL5S8Z0ssEKcDBGMTANNnOVQnBJZXj6PUVKqa+y2E48MQ==
X-MS-Exchange-Transport-CrossTenantHeadersStamped: CH2PR22MB1847
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/MsW6qvFD3Q5zbMXVstQfaKCKdMA>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 30 Jun 2020 16:16:53 -0000

--_000_CH2PR22MB20869C49226FFDD90CED291DDA6F0CH2PR22MB2086namp_
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

I can see where that output would be confusing.  This is mandated by the sp=
ec in Section 8.1, which says:


Clients MUST ensure that UDP datagrams containing Initial packets have UDP =
payloads of at least 1200 bytes, adding padding to packets in the datagram =
as necessary. A client that sends padded datagrams allows the server to sen=
d more data prior to completing address validation.

Because the client's address has not been validated when the server sends i=
ts initial response, there's an obvious risk of an amplification attack if =
the server's response is large.  To mitigate this, QUIC requires that clien=
ts send large Initial packets and restricts the bytes servers send to unver=
ified addresses to a fixed multiple of the bytes received from the client. =
 Only UDP datagrams containing Initial packets have this requirement, becau=
se once you've moved to other packet types, the address has been validated.

From: QUIC <quic-bounces@ietf.org> On Behalf Of tim.jebb@bt.com
Sent: Tuesday, June 30, 2020 6:16 AM
To: quic@ietf.org
Subject: Padding in Initial Packets

Hi All,

I'm fairly new to QUIC so apologies if the questions in here are dumb.

I've been going through a QUIC pcap, which I captured from a Firefox implem=
entation of IETF draft 27 QUIC, and am a little confused about the use of p=
adding in the first Initial packet.  According to Wireshark,  the UDP datag=
ram contains a QUIC packet of type initial, length 272, the payload of whic=
h is a Client Hello.

The remainder of the UDP datagram is 1055 zero bytes..  According to Wiresh=
ark these zero bytes constitute a QUIC IETF Packet.

However there doesn't seem to be a valid QUIC packet header for these bytes=
.  AIUI according to the QUIC spec padding bytes are frames within a packet=
, so these zero bytes seem to be just bytes in the UDP datagram outside of =
the QUIC payload.  Which would be fine as far as I can see, but:

Why does Wireshark say it's QUIC IETF?
Is there anything in any of the QUIC draft RFCs mandating this behaviour, i=
f so where?
Why is it being done - can't see the point?
If padding is to be added surely it should be as part of a QUIC packet and =
encrypted - here there's just a load of plaintext zero bytes.


Here is Wireshark's analysis:

"QUIC IETF
    QUIC Connection information
    [Packet Length: 302]
    1... .... =3D Header Form: Long Header (1)
    .1.. .... =3D Fixed Bit: True
    ..00 .... =3D Packet Type: Initial (0)
    .... 00.. =3D Reserved: 0
    .... ..10 =3D Packet Number Length: 3 bytes (2)
    Version: draft-27 (0xff00001b)
    Destination Connection ID Length: 17
    Destination Connection ID: 76f81a5588b2f25c5ab6dbd461a8d749e1
    Source Connection ID Length: 3
    Source Connection ID: c5ef02
    Token Length: 0
    Length: 272
    Packet Number: 0
    Payload: 04f535177fcbcfd5f06d24ffde66edf10df79205666b1851...
    TLSv1.3 Record Layer: Handshake Protocol: Client Hello
        Frame Type: CRYPTO (0x0000000000000006)
        Offset: 0
        Length: 249
        Crypto Data
        Handshake Protocol: Client Hello
            Handshake Type: Client Hello (1)
            Length: 245
            Version: TLS 1.2 (0x0303)
            Random: b1b76126ddab3c639cdfaa7d92d6cabe2ea9c1a19ea86b8b...
            Session ID Length: 0
            Cipher Suites Length: 4
            Cipher Suites (2 suites)
            Compression Methods Length: 1
            Compression Methods (1 method)
            Extensions Length: 200
            Extension: server_name (len=3D15)
            Extension: extended_master_secret (len=3D0)
            Extension: renegotiation_info (len=3D1)
            Extension: supported_groups (len=3D20)
            Extension: application_layer_protocol_negotiation (len=3D8)
            Extension: status_request (len=3D5)
            Extension: key_share (len=3D38)
            Extension: supported_versions (len=3D3)
            Extension: signature_algorithms (len=3D16)
            Extension: psk_key_exchange_modes (len=3D2)
            Extension: record_size_limit (len=3D2)
            Extension: quic_transports_parameters (len=3D42)
QUIC IETF
    [Packet Length: 1055]
    Remaining Payload: 000000000000000000000000000000000000000000000000..."

Any comments or explanations?

Tim

--_000_CH2PR22MB20869C49226FFDD90CED291DDA6F0CH2PR22MB2086namp_
Content-Type: text/html; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

<html xmlns:v=3D"urn:schemas-microsoft-com:vml" xmlns:o=3D"urn:schemas-micr=
osoft-com:office:office" xmlns:w=3D"urn:schemas-microsoft-com:office:word" =
xmlns:m=3D"http://schemas.microsoft.com/office/2004/12/omml" xmlns=3D"http:=
//www.w3.org/TR/REC-html40">
<head>
<meta http-equiv=3D"Content-Type" content=3D"text/html; charset=3Dus-ascii"=
>
<meta name=3D"Generator" content=3D"Microsoft Word 15 (filtered medium)">
<style><!--
/* Font Definitions */
@font-face
	{font-family:Wingdings;
	panose-1:5 0 0 0 0 0 0 0 0 0;}
@font-face
	{font-family:"Cambria Math";
	panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
	{font-family:Calibri;
	panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
	{margin:0in;
	margin-bottom:.0001pt;
	font-size:11.0pt;
	font-family:"Calibri",sans-serif;}
p.MsoListParagraph, li.MsoListParagraph, div.MsoListParagraph
	{mso-style-priority:34;
	margin-top:0in;
	margin-right:0in;
	margin-bottom:0in;
	margin-left:.5in;
	margin-bottom:.0001pt;
	font-size:11.0pt;
	font-family:"Calibri",sans-serif;}
.MsoChpDefault
	{mso-style-type:export-only;
	font-size:10.0pt;}
@page WordSection1
	{size:8.5in 11.0in;
	margin:1.0in 1.0in 1.0in 1.0in;}
div.WordSection1
	{page:WordSection1;}
/* List Definitions */
@list l0
	{mso-list-id:380592662;
	mso-list-type:hybrid;
	mso-list-template-ids:-787426934 -2007181428 67698691 67698693 67698689 67=
698691 67698693 67698689 67698691 67698693;}
@list l0:level1
	{mso-level-start-at:0;
	mso-level-number-format:bullet;
	mso-level-text:\F0D8;
	mso-level-tab-stop:none;
	mso-level-number-position:left;
	text-indent:-.25in;
	font-family:Wingdings;
	mso-fareast-font-family:Calibri;
	mso-bidi-font-family:"Times New Roman";}
@list l0:level2
	{mso-level-number-format:bullet;
	mso-level-text:o;
	mso-level-tab-stop:none;
	mso-level-number-position:left;
	text-indent:-.25in;
	font-family:"Courier New";}
@list l0:level3
	{mso-level-number-format:bullet;
	mso-level-text:\F0A7;
	mso-level-tab-stop:none;
	mso-level-number-position:left;
	text-indent:-.25in;
	font-family:Wingdings;}
@list l0:level4
	{mso-level-number-format:bullet;
	mso-level-text:\F0B7;
	mso-level-tab-stop:none;
	mso-level-number-position:left;
	text-indent:-.25in;
	font-family:Symbol;}
@list l0:level5
	{mso-level-number-format:bullet;
	mso-level-text:o;
	mso-level-tab-stop:none;
	mso-level-number-position:left;
	text-indent:-.25in;
	font-family:"Courier New";}
@list l0:level6
	{mso-level-number-format:bullet;
	mso-level-text:\F0A7;
	mso-level-tab-stop:none;
	mso-level-number-position:left;
	text-indent:-.25in;
	font-family:Wingdings;}
@list l0:level7
	{mso-level-number-format:bullet;
	mso-level-text:\F0B7;
	mso-level-tab-stop:none;
	mso-level-number-position:left;
	text-indent:-.25in;
	font-family:Symbol;}
@list l0:level8
	{mso-level-number-format:bullet;
	mso-level-text:o;
	mso-level-tab-stop:none;
	mso-level-number-position:left;
	text-indent:-.25in;
	font-family:"Courier New";}
@list l0:level9
	{mso-level-number-format:bullet;
	mso-level-text:\F0A7;
	mso-level-tab-stop:none;
	mso-level-number-position:left;
	text-indent:-.25in;
	font-family:Wingdings;}
ol
	{margin-bottom:0in;}
ul
	{margin-bottom:0in;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext=3D"edit" spidmax=3D"1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext=3D"edit">
<o:idmap v:ext=3D"edit" data=3D"1" />
</o:shapelayout></xml><![endif]-->
</head>
<body lang=3D"EN-US" link=3D"#0563C1" vlink=3D"#954F72">
<div class=3D"WordSection1">
<p class=3D"MsoNormal">I can see where that output would be confusing.&nbsp=
; This is mandated by the spec in Section 8.1, which says:<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoListParagraph"><span style=3D"font-size:12.0pt;font-family:&=
quot;Times New Roman&quot;,serif;color:#222222;background:white">Clients MU=
ST ensure that UDP datagrams containing Initial packets have UDP payloads o=
f at least 1200 bytes, adding padding to packets
 in the datagram as necessary. A client that sends padded datagrams allows =
the server to send more data prior to completing address validation.</span>=
<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">Because the client&#8217;s address has not been vali=
dated when the server sends its initial response, there&#8217;s an obvious =
risk of an amplification attack if the server&#8217;s response is large.&nb=
sp; To mitigate this, QUIC requires that clients send large
 Initial packets and restricts the bytes servers send to unverified address=
es to a fixed multiple of the bytes received from the client.&nbsp; Only UD=
P datagrams containing Initial packets have this requirement, because once =
you&#8217;ve moved to other packet types, the
 address has been validated.<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<div>
<div style=3D"border:none;border-top:solid #E1E1E1 1.0pt;padding:3.0pt 0in =
0in 0in">
<p class=3D"MsoNormal"><b>From:</b> QUIC &lt;quic-bounces@ietf.org&gt; <b>O=
n Behalf Of </b>
tim.jebb@bt.com<br>
<b>Sent:</b> Tuesday, June 30, 2020 6:16 AM<br>
<b>To:</b> quic@ietf.org<br>
<b>Subject:</b> Padding in Initial Packets<o:p></o:p></p>
</div>
</div>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">Hi All,<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB"><o:p>&nbsp;</o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">I&#8217;m fairly new to QUIC so=
 apologies if the questions in here are dumb.<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB"><o:p>&nbsp;</o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">I&#8217;ve been going through a=
 QUIC pcap, which I captured from a Firefox implementation of IETF draft 27=
 QUIC, and am a little confused about the use of padding in the first Initi=
al packet.&nbsp; According to Wireshark,&nbsp; the UDP
 datagram contains a QUIC packet of type initial, length 272, the payload o=
f which is a Client Hello.&nbsp;
<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB"><o:p>&nbsp;</o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">The remainder of the UDP datagr=
am is 1055 zero bytes..&nbsp; According to Wireshark these zero bytes const=
itute a QUIC IETF Packet.<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB"><o:p>&nbsp;</o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">However there doesn&#8217;t see=
m to be a valid QUIC packet header for these bytes.&nbsp; AIUI according to=
 the QUIC spec padding bytes are frames within a packet, so these zero byte=
s seem to be just bytes in the UDP datagram outside
 of the QUIC payload.&nbsp; Which would be fine as far as I can see, but:<o=
:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB"><o:p>&nbsp;</o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">Why does Wireshark say it&#8217=
;s QUIC IETF?<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">Is there anything in any of the=
 QUIC draft RFCs mandating this behaviour, if so where?<o:p></o:p></span></=
p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">Why is it being done &#8211; ca=
n&#8217;t see the point?<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">If padding is to be added surel=
y it should be as part of a QUIC packet and encrypted &#8211; here there&#8=
217;s just a load of plaintext zero bytes.<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB"><o:p>&nbsp;</o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB"><o:p>&nbsp;</o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">Here is Wireshark&#8217;s analy=
sis:<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB"><o:p>&nbsp;</o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&#8220;QUIC IETF<o:p></o:p></sp=
an></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; QUIC Connect=
ion information<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; [Packet Leng=
th: 302]<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; 1... .... =
=3D Header Form: Long Header (1)<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; .1.. .... =
=3D Fixed Bit: True<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; ..00 .... =
=3D Packet Type: Initial (0)<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; .... 00.. =
=3D Reserved: 0<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; .... ..10 =
=3D Packet Number Length: 3 bytes (2)<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; Version: dra=
ft-27 (0xff00001b)<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; Destination =
Connection ID Length: 17<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; Destination =
Connection ID: 76f81a5588b2f25c5ab6dbd461a8d749e1<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; Source Conne=
ction ID Length: 3<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; Source Conne=
ction ID: c5ef02<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; Token Length=
: 0<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; Length: 272<=
o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; Packet Numbe=
r: 0<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; Payload: 04f=
535177fcbcfd5f06d24ffde66edf10df79205666b1851&#8230;<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; TLSv1.3 Reco=
rd Layer: Handshake Protocol: Client Hello<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp; Frame Type: CRYPTO (0x0000000000000006)<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp; Offset: 0<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp; Length: 249<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp; Crypto Data<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp; Handshake Protocol: Client Hello<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Handshake Type: Client Hello (1)<o:p></=
o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Length: 245<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Version: TLS 1.2 (0x0303)<o:p></o:p></s=
pan></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Random: b1b76126ddab3c639cdfaa7d92d6cab=
e2ea9c1a19ea86b8b&#8230;<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Session ID Length: 0<o:p></o:p></span><=
/p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Cipher Suites Length: 4<o:p></o:p></spa=
n></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Cipher Suites (2 suites)<o:p></o:p></sp=
an></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Compression Methods Length: 1<o:p></o:p=
></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Compression Methods (1 method)<o:p></o:=
p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Extensions Length: 200<o:p></o:p></span=
></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Extension: server_name (len=3D15)<o:p><=
/o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Extension: extended_master_secret (len=
=3D0)<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Extension: renegotiation_info (len=3D1)=
<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Extension: supported_groups (len=3D20)<=
o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Extension: application_layer_protocol_n=
egotiation (len=3D8)<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Extension: status_request (len=3D5)<o:p=
></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Extension: key_share (len=3D38)<o:p></o=
:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Extension: supported_versions (len=3D3)=
<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Extension: signature_algorithms (len=3D=
16)<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Extension: psk_key_exchange_modes (len=
=3D2)<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Extension: record_size_limit (len=3D2)<=
o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Extension: quic_transports_parameters (=
len=3D42)<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">QUIC IETF<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; [Packet Leng=
th: 1055]<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">&nbsp;&nbsp;&nbsp; Remaining Pa=
yload: 000000000000000000000000000000000000000000000000&#8230;&#8221;<o:p><=
/o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB"><o:p>&nbsp;</o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">Any comments or explanations?<o=
:p></o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB"><o:p>&nbsp;</o:p></span></p>
<p class=3D"MsoNormal"><span lang=3D"EN-GB">Tim<o:p></o:p></span></p>
</div>
</body>
</html>

--_000_CH2PR22MB20869C49226FFDD90CED291DDA6F0CH2PR22MB2086namp_--


From nobody Tue Jun 30 18:18:51 2020
Return-Path: <mt@lowentropy.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 03C0A3A086B for <quic@ietfa.amsl.com>; Tue, 30 Jun 2020 18:18:50 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.098
X-Spam-Level: 
X-Spam-Status: No, score=-2.098 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_MSPIKE_H3=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=lowentropy.net header.b=numkcfAW; dkim=pass (2048-bit key) header.d=messagingengine.com header.b=trk8f9N4
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id awuT_yr_qYXC for <quic@ietfa.amsl.com>; Tue, 30 Jun 2020 18:18:48 -0700 (PDT)
Received: from out5-smtp.messagingengine.com (out5-smtp.messagingengine.com [66.111.4.29]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 22C403A086D for <quic@ietf.org>; Tue, 30 Jun 2020 18:18:48 -0700 (PDT)
Received: from compute2.internal (compute2.nyi.internal [10.202.2.42]) by mailout.nyi.internal (Postfix) with ESMTP id 12FE15C0116 for <quic@ietf.org>; Tue, 30 Jun 2020 21:18:47 -0400 (EDT)
Received: from imap2 ([10.202.2.52]) by compute2.internal (MEProxy); Tue, 30 Jun 2020 21:18:47 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=lowentropy.net; h=mime-version:message-id:in-reply-to:references:date:from:to :subject:content-type:content-transfer-encoding; s=fm2; bh=/sh5e asAr4mExRXrA4kCZrKQm/C0PrGRI9KWWxXj2H8=; b=numkcfAWWLdHuoRMEbjlB LzdGKQIZ5R997dDIEUtIqEDJGiIYXHSyJYqKeQKAu9wMvVzAHY2i5UR86PYBkeqB NL52PQ5rqD15ts61oMrYaftPJFjTOEN0GUhRVPfygJz3JI6CQBf835TMnb4hogtd RR+0Tz6zOjjBrcawvM8vseiOcOaWjUnlKFp5M0JWPlwbijBrENU3RM7WW+Orcm/L iOpioNTMXpCzEYpgODa0KVROvqo7kSA1OUrzgBcOzlWOvM+x+DjMcnEbMHkjvy/L SCYZgawKT6m/EZAE6XPvw/s/9Ld2rp0B56ba+JVr+UM9/G8XyxyPSHll8j9ez/K9 w==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d= messagingengine.com; h=content-transfer-encoding:content-type :date:from:in-reply-to:message-id:mime-version:references :subject:to:x-me-proxy:x-me-proxy:x-me-sender:x-me-sender :x-sasl-enc; s=fm3; bh=/sh5easAr4mExRXrA4kCZrKQm/C0PrGRI9KWWxXj2 H8=; b=trk8f9N4FCTBvNwS6Bc6Ko2bEtg9jZVEPzLtuyq2S+FAwhSKHdwLc08Uk 0z5eno32cQQis58fWYTipkwDSKl7FJCIrLsFx9ng7UzvW0fxqFtJDRlmJpNZv80e qeTL/HjQC7C2t6a4vSDSQPdbs1G1iFM+cQcD/DJ4+g7+K1xkRZpxzd7lwLg+PvI9 gy4fNH5IGDp66ig7htWFLfowt9cIhWryR7THOcE9jIMUZta3Jreyk4uGL1AehRYi 03RoIJanVG3j2HfXhYsYD294TIK//Lt3yoLeUPgT4YS5McHd60y/1w5eWoyR0PQ2 M7dexj9b/XR6sQ1bTahsKil/+eGSg==
X-ME-Sender: <xms:duT7XoqjOoQz_om_LpOtDOavy3dw9Q5vUf9tgh_BVxUtgcNS96cMig>
X-ME-Proxy-Cause: gggruggvucftvghtrhhoucdtuddrgeduiedrtddugdeghecutefuodetggdotefrodftvf curfhrohhfihhlvgemucfhrghsthforghilhdpqfgfvfdpuffrtefokffrpgfnqfghnecu uegrihhlohhuthemuceftddtnecunecujfgurhepofgfggfkjghffffhvffutgfgsehtqh ertderreejnecuhfhrohhmpedfofgrrhhtihhnucfvhhhomhhsohhnfdcuoehmtheslhho figvnhhtrhhophihrdhnvghtqeenucggtffrrghtthgvrhhnpefgjeeuudeiffeltdegle ehjedvtedtjeduudehgfegfefgkefgueeiteegffdttdenucevlhhushhtvghrufhiiigv pedtnecurfgrrhgrmhepmhgrihhlfhhrohhmpehmtheslhhofigvnhhtrhhophihrdhnvg ht
X-ME-Proxy: <xmx:duT7Xuq5E7N9awGmeT1dmiAfEWOY34hHb4UAHzlMN6VrYGzKLdmcGQ> <xmx:duT7XtO8hj8lMD96SbVsevuDkvEcu9YvOXy0yszmfJ4JfbkCtAmBmg> <xmx:duT7Xv5ja2vrnfHUQtxhPEusyXW1BttAy3RbNcqeY6JUlRryBTJCPQ> <xmx:d-T7XqKQEUkfi3A2BR4nMJ2ip4da3Rc3KFvTtM0EHxOKi0zEj9UP9g>
Received: by mailuser.nyi.internal (Postfix, from userid 501) id 56F92E00CE; Tue, 30 Jun 2020 21:18:46 -0400 (EDT)
X-Mailer: MessagingEngine.com Webmail Interface
User-Agent: Cyrus-JMAP/3.3.0-dev0-576-gfe2cd66-fm-20200629.001-gfe2cd668
Mime-Version: 1.0
Message-Id: <f97dbbe6-e719-4276-bd19-da6589bb2d4e@www.fastmail.com>
In-Reply-To: <LNXP123MB24602880B0AE644B9B1F3F83F26F0@LNXP123MB2460.GBRP123.PROD.OUTLOOK.COM>
References: <LNXP123MB24602880B0AE644B9B1F3F83F26F0@LNXP123MB2460.GBRP123.PROD.OUTLOOK.COM>
Date: Wed, 01 Jul 2020 11:18:05 +1000
From: "Martin Thomson" <mt@lowentropy.net>
To: quic@ietf.org
Subject: Re: Padding in Initial Packets
Content-Type: text/plain;charset=utf-8
Content-Transfer-Encoding: quoted-printable
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/v0kO1KxIyF4x_RC2_2tuXRcRPc4>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 01 Jul 2020 01:18:50 -0000

Hi Tim,

That's a little curveball that Firefox throws.  Other implementations pa=
d with PADDING frames inside Initial packets.  Our first implementation =
found that inconvenient and so we padded outside the packet.  The result=
ing "packets" are clearly not valid, but neither server nor spec cannot =
require that they be legitimate. QUIC doesn't authenticate the compositi=
on of QUIC packets into UDP datagrams.  So the overall result is valid, =
even if the individual bits are not.

The number of bugs this behaviour has exposed is so great that we ultima=
tely decided to keep it.  Fixing it would make our code more complicated=
 anyway.

We might change the value we use for padding from 0 to something else at=
 some point, just to keep things fresh.

Cheers,
Martin

On Tue, Jun 30, 2020, at 20:15, tim.jebb@bt.com wrote:
> =20
> Hi All,
>=20
>=20
> I=E2=80=99m fairly new to QUIC so apologies if the questions in here a=
re dumb.
>=20
>=20
> I=E2=80=99ve been going through a QUIC pcap, which I captured from a F=
irefox=20
> implementation of IETF draft 27 QUIC, and am a little confused about=20=

> the use of padding in the first Initial packet. According to Wireshark=
,=20
> the UDP datagram contains a QUIC packet of type initial, length 272,=20=

> the payload of which is a Client Hello.=20
>=20
>=20
> The remainder of the UDP datagram is 1055 zero bytes.. According to=20=

> Wireshark these zero bytes constitute a QUIC IETF Packet.
>=20
>=20
> However there doesn=E2=80=99t seem to be a valid QUIC packet header fo=
r these=20
> bytes. AIUI according to the QUIC spec padding bytes are frames within=
=20
> a packet, so these zero bytes seem to be just bytes in the UDP datagra=
m=20
> outside of the QUIC payload. Which would be fine as far as I can see,=20=

> but:
>=20
>=20
> Why does Wireshark say it=E2=80=99s QUIC IETF?
>=20
> Is there anything in any of the QUIC draft RFCs mandating this=20
> behaviour, if so where?
>=20
> Why is it being done =E2=80=93 can=E2=80=99t see the point?
>=20
> If padding is to be added surely it should be as part of a QUIC packet=
=20
> and encrypted =E2=80=93 here there=E2=80=99s just a load of plaintext =
zero bytes.
>=20
>=20
>=20
> Here is Wireshark=E2=80=99s analysis:
>=20
>=20
> =E2=80=9CQUIC IETF
>=20
>  QUIC Connection information
>=20
>  [Packet Length: 302]
>=20
>  1... .... =3D Header Form: Long Header (1)
>=20
>  .1.. .... =3D Fixed Bit: True
>=20
>  ..00 .... =3D Packet Type: Initial (0)
>=20
>  .... 00.. =3D Reserved: 0
>=20
>  .... ..10 =3D Packet Number Length: 3 bytes (2)
>=20
>  Version: draft-27 (0xff00001b)
>=20
>  Destination Connection ID Length: 17
>=20
>  Destination Connection ID: 76f81a5588b2f25c5ab6dbd461a8d749e1
>=20
>  Source Connection ID Length: 3
>=20
>  Source Connection ID: c5ef02
>=20
>  Token Length: 0
>=20
>  Length: 272
>=20
>  Packet Number: 0
>=20
>  Payload: 04f535177fcbcfd5f06d24ffde66edf10df79205666b1851=E2=80=A6
>=20
>  TLSv1.3 Record Layer: Handshake Protocol: Client Hello
>=20
>  Frame Type: CRYPTO (0x0000000000000006)
>=20
>  Offset: 0
>=20
>  Length: 249
>=20
>  Crypto Data
>=20
>  Handshake Protocol: Client Hello
>=20
>  Handshake Type: Client Hello (1)
>=20
>  Length: 245
>=20
>  Version: TLS 1.2 (0x0303)
>=20
>  Random: b1b76126ddab3c639cdfaa7d92d6cabe2ea9c1a19ea86b8b=E2=80=A6
>=20
>  Session ID Length: 0
>=20
>  Cipher Suites Length: 4
>=20
>  Cipher Suites (2 suites)
>=20
>  Compression Methods Length: 1
>=20
>  Compression Methods (1 method)
>=20
>  Extensions Length: 200
>=20
>  Extension: server_name (len=3D15)
>=20
>  Extension: extended_master_secret (len=3D0)
>=20
>  Extension: renegotiation_info (len=3D1)
>=20
>  Extension: supported_groups (len=3D20)
>=20
>  Extension: application_layer_protocol_negotiation (len=3D8)
>=20
>  Extension: status_request (len=3D5)
>=20
>  Extension: key_share (len=3D38)
>=20
>  Extension: supported_versions (len=3D3)
>=20
>  Extension: signature_algorithms (len=3D16)
>=20
>  Extension: psk_key_exchange_modes (len=3D2)
>=20
>  Extension: record_size_limit (len=3D2)
>=20
>  Extension: quic_transports_parameters (len=3D42)
>=20
> QUIC IETF
>=20
>  [Packet Length: 1055]
>=20
>  Remaining Payload: 000000000000000000000000000000000000000000000000=E2=
=80=A6=E2=80=9D
>=20
>=20
> Any comments or explanations?
>=20
>=20
> Tim
>


From nobody Tue Jun 30 19:17:03 2020
Return-Path: <lars@eggert.org>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 478C03A090D for <quic@ietfa.amsl.com>; Tue, 30 Jun 2020 19:17:02 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.099
X-Spam-Level: 
X-Spam-Status: No, score=-2.099 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=eggert.org
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id jpD7oIBX_pQg for <quic@ietfa.amsl.com>; Tue, 30 Jun 2020 19:17:00 -0700 (PDT)
Received: from mail.eggert.org (mail.eggert.org [IPv6:2a00:ac00:0:35:211:32ff:fe22:186f]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 76C713A08F8 for <quic@ietf.org>; Tue, 30 Jun 2020 19:17:00 -0700 (PDT)
Received: from [IPv6:2a00:ac00:0:35:881:7bbf:55a7:1186] (unknown [IPv6:2a00:ac00:0:35:881:7bbf:55a7:1186]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.eggert.org (Postfix) with ESMTPSA id 86B06611683; Wed,  1 Jul 2020 05:16:53 +0300 (EEST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=eggert.org; s=dkim; t=1593569813; bh=TS5HsPWpkrRXtO14h5eyMAmWpyPIlTMuumoJiNPUNEU=; h=From:Subject:Date:References:Cc:In-Reply-To:To; b=wbbzVcFr1VyECrMr1GXajqKEGcAGQmeB1BjEHq3AixY1LLnyOz8Va7mXe/mdz3jEk 0eJIl5JpYKMey7uXMpQcqekHLypYB6ns6RwkQNzexhGKEIAXoZEX1GXVB6sQYTGXu5 S/AwK3uXfJvrfzmsgbut8piNitIeuMZUSkOVXQyM=
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
From: Lars Eggert <lars@eggert.org>
Mime-Version: 1.0 (1.0)
Subject: Re: Padding in Initial Packets
Date: Wed, 1 Jul 2020 05:16:50 +0300
Message-Id: <E86FC27B-E38E-4842-BD4F-A316ADD51FE9@eggert.org>
References: <f97dbbe6-e719-4276-bd19-da6589bb2d4e@www.fastmail.com>
Cc: quic@ietf.org
In-Reply-To: <f97dbbe6-e719-4276-bd19-da6589bb2d4e@www.fastmail.com>
To: Martin Thomson <mt@lowentropy.net>
X-MailScanner-ID: 86B06611683.A14BF
X-MailScanner: Found to be clean
X-MailScanner-From: lars@eggert.org
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/aHtLj79Xg8niNFNSFmUuvQImRRg>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 01 Jul 2020 02:17:02 -0000

Quant similarly pads by coalescing random data (with the first byte massaged=
 to be of short-header type).

--=20
Sent from a mobile device; please excuse typos.

> On Jul 1, 2020, at 04:19, Martin Thomson <mt@lowentropy.net> wrote:
>=20
> =EF=BB=BFHi Tim,
>=20
> That's a little curveball that Firefox throws.  Other implementations pad w=
ith PADDING frames inside Initial packets.  Our first implementation found t=
hat inconvenient and so we padded outside the packet.  The resulting "packet=
s" are clearly not valid, but neither server nor spec cannot require that th=
ey be legitimate. QUIC doesn't authenticate the composition of QUIC packets i=
nto UDP datagrams.  So the overall result is valid, even if the individual b=
its are not.
>=20
> The number of bugs this behaviour has exposed is so great that we ultimate=
ly decided to keep it.  Fixing it would make our code more complicated anywa=
y.
>=20
> We might change the value we use for padding from 0 to something else at s=
ome point, just to keep things fresh.
>=20
> Cheers,
> Martin
>=20
>> On Tue, Jun 30, 2020, at 20:15, tim.jebb@bt.com wrote:
>>=20
>> Hi All,
>>=20
>>=20
>> I=E2=80=99m fairly new to QUIC so apologies if the questions in here are d=
umb.
>>=20
>>=20
>> I=E2=80=99ve been going through a QUIC pcap, which I captured from a Fire=
fox=20
>> implementation of IETF draft 27 QUIC, and am a little confused about=20
>> the use of padding in the first Initial packet. According to Wireshark,=20=

>> the UDP datagram contains a QUIC packet of type initial, length 272,=20
>> the payload of which is a Client Hello.=20
>>=20
>>=20
>> The remainder of the UDP datagram is 1055 zero bytes.. According to=20
>> Wireshark these zero bytes constitute a QUIC IETF Packet.
>>=20
>>=20
>> However there doesn=E2=80=99t seem to be a valid QUIC packet header for t=
hese=20
>> bytes. AIUI according to the QUIC spec padding bytes are frames within=20=

>> a packet, so these zero bytes seem to be just bytes in the UDP datagram=20=

>> outside of the QUIC payload. Which would be fine as far as I can see,=20
>> but:
>>=20
>>=20
>> Why does Wireshark say it=E2=80=99s QUIC IETF?
>>=20
>> Is there anything in any of the QUIC draft RFCs mandating this=20
>> behaviour, if so where?
>>=20
>> Why is it being done =E2=80=93 can=E2=80=99t see the point?
>>=20
>> If padding is to be added surely it should be as part of a QUIC packet=20=

>> and encrypted =E2=80=93 here there=E2=80=99s just a load of plaintext zer=
o bytes.
>>=20
>>=20
>>=20
>> Here is Wireshark=E2=80=99s analysis:
>>=20
>>=20
>> =E2=80=9CQUIC IETF
>>=20
>> QUIC Connection information
>>=20
>> [Packet Length: 302]
>>=20
>> 1... .... =3D Header Form: Long Header (1)
>>=20
>> .1.. .... =3D Fixed Bit: True
>>=20
>> ..00 .... =3D Packet Type: Initial (0)
>>=20
>> .... 00.. =3D Reserved: 0
>>=20
>> .... ..10 =3D Packet Number Length: 3 bytes (2)
>>=20
>> Version: draft-27 (0xff00001b)
>>=20
>> Destination Connection ID Length: 17
>>=20
>> Destination Connection ID: 76f81a5588b2f25c5ab6dbd461a8d749e1
>>=20
>> Source Connection ID Length: 3
>>=20
>> Source Connection ID: c5ef02
>>=20
>> Token Length: 0
>>=20
>> Length: 272
>>=20
>> Packet Number: 0
>>=20
>> Payload: 04f535177fcbcfd5f06d24ffde66edf10df79205666b1851=E2=80=A6
>>=20
>> TLSv1.3 Record Layer: Handshake Protocol: Client Hello
>>=20
>> Frame Type: CRYPTO (0x0000000000000006)
>>=20
>> Offset: 0
>>=20
>> Length: 249
>>=20
>> Crypto Data
>>=20
>> Handshake Protocol: Client Hello
>>=20
>> Handshake Type: Client Hello (1)
>>=20
>> Length: 245
>>=20
>> Version: TLS 1.2 (0x0303)
>>=20
>> Random: b1b76126ddab3c639cdfaa7d92d6cabe2ea9c1a19ea86b8b=E2=80=A6
>>=20
>> Session ID Length: 0
>>=20
>> Cipher Suites Length: 4
>>=20
>> Cipher Suites (2 suites)
>>=20
>> Compression Methods Length: 1
>>=20
>> Compression Methods (1 method)
>>=20
>> Extensions Length: 200
>>=20
>> Extension: server_name (len=3D15)
>>=20
>> Extension: extended_master_secret (len=3D0)
>>=20
>> Extension: renegotiation_info (len=3D1)
>>=20
>> Extension: supported_groups (len=3D20)
>>=20
>> Extension: application_layer_protocol_negotiation (len=3D8)
>>=20
>> Extension: status_request (len=3D5)
>>=20
>> Extension: key_share (len=3D38)
>>=20
>> Extension: supported_versions (len=3D3)
>>=20
>> Extension: signature_algorithms (len=3D16)
>>=20
>> Extension: psk_key_exchange_modes (len=3D2)
>>=20
>> Extension: record_size_limit (len=3D2)
>>=20
>> Extension: quic_transports_parameters (len=3D42)
>>=20
>> QUIC IETF
>>=20
>> [Packet Length: 1055]
>>=20
>> Remaining Payload: 000000000000000000000000000000000000000000000000=E2=80=
=A6=E2=80=9D
>>=20
>>=20
>> Any comments or explanations?
>>=20
>>=20
>> Tim
>>=20
>=20


From nobody Tue Jun 30 19:33:46 2020
Return-Path: <mt@lowentropy.net>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 05E133A094E for <quic@ietfa.amsl.com>; Tue, 30 Jun 2020 19:33:45 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.098
X-Spam-Level: 
X-Spam-Status: No, score=-2.098 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_MSPIKE_H3=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=lowentropy.net header.b=X6QUIOjl; dkim=pass (2048-bit key) header.d=messagingengine.com header.b=XVj3gncP
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id sA7UkZHKqAgQ for <quic@ietfa.amsl.com>; Tue, 30 Jun 2020 19:33:43 -0700 (PDT)
Received: from out5-smtp.messagingengine.com (out5-smtp.messagingengine.com [66.111.4.29]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id A7BE03A0941 for <quic@ietf.org>; Tue, 30 Jun 2020 19:33:43 -0700 (PDT)
Received: from compute2.internal (compute2.nyi.internal [10.202.2.42]) by mailout.nyi.internal (Postfix) with ESMTP id E73B45C0108 for <quic@ietf.org>; Tue, 30 Jun 2020 22:33:42 -0400 (EDT)
Received: from imap2 ([10.202.2.52]) by compute2.internal (MEProxy); Tue, 30 Jun 2020 22:33:42 -0400
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=lowentropy.net; h=mime-version:message-id:date:from:to:subject:content-type; s= fm2; bh=VxxB1w8n5dRmq9w2wauY9bDonRUNeuf+Vg2PjRl3TLM=; b=X6QUIOjl eOZNXMOf2s2MQWflvroX6pUOQU/1rPbdTQ1pbbMlB+gmumxflJ1ClTCzvSSHDcif 3/Rl4lsoKO7Q9RpREjHJIZsbY1GElcNCU9DlQkVSh3L72fArplSeD/zW5ROjmnTt ChHNPI4gLQmRAAsaQKr1V48df+bxHpZo8f7GzUDdMXKGFYM8Gt4YcFhoyv7h/51O cLdYAHDK2Spvrn1EEtnEpQ7arP960WbjQj+i+ukhen9pnSQr/c4PE9MPH2R/b2yx qrZ91NkQTb2aAIx12rT9ZLPiVopPSVZTxnQkf0JfngEXor77RaAE83qInCk2bCIq ltq/sZLIuJkwtw==
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d= messagingengine.com; h=content-type:date:from:message-id :mime-version:subject:to:x-me-proxy:x-me-proxy:x-me-sender :x-me-sender:x-sasl-enc; s=fm3; bh=VxxB1w8n5dRmq9w2wauY9bDonRUNe uf+Vg2PjRl3TLM=; b=XVj3gncPM8NjN6Q0gcLq7JsFrTaUsXtZBs49g/bigZ8pg OXdzmmV/8bbGuRwkQTWEzYVMY+Bk8rsiukwvC1R/b7VzWPyE01KUFIwxuyE9kkOX PIsa+qouCrl2yQjX9waDP3NIECRlMNEFAIdvJcN/cTX1eWVK9WtUT8nlZqRa3i8f YkdWeD7TWY9Xf3AIAVYpurJcqpuACPA1BQIew/fvwV3LYCUA5877RQ8TO0xSHLgX Y1y5TxFUxG6RbGLQZCt25/Iu8SRvjMmnDsUmggjSecdhyYPlDYhwnC05vJwog/Uk DIQeqXSNjaDhcR9wPWNUPse1d2QF6Hk3zNnBBSlMg==
X-ME-Sender: <xms:Bvb7Xpx2MF964-mrtSncT2Xp7vViC3gx0kxH8Ocj6Andke7HCjuWug>
X-ME-Proxy-Cause: gggruggvucftvghtrhhoucdtuddrgeduiedrtddugdeivdcutefuodetggdotefrodftvf curfhrohhfihhlvgemucfhrghsthforghilhdpqfgfvfdpuffrtefokffrpgfnqfghnecu uegrihhlohhuthemuceftddtnecunecujfgurhepofgfggfkfffhvffutgesthdtredtre ertdenucfhrhhomhepfdforghrthhinhcuvfhhohhmshhonhdfuceomhhtsehlohifvghn thhrohhphidrnhgvtheqnecuggftrfgrthhtvghrnhepgeeuheeujedvieevffehvdekge ffkefggfdtteelfeekfeeftdejhfdtuddvuddtnecuffhomhgrihhnpehgihhthhhusgdr tghomhenucevlhhushhtvghrufhiiigvpedtnecurfgrrhgrmhepmhgrihhlfhhrohhmpe hmtheslhhofigvnhhtrhhophihrdhnvght
X-ME-Proxy: <xmx:Bvb7XpTE-Pm344JQ4roIZDBjXEcOKpKjQRefieuzrGRfgailQ136EQ> <xmx:Bvb7XjUDWXdWbutxzE6MiDIqdD____UN4VjLG--VAUCFGEd88QAtyg> <xmx:Bvb7Xrhnh_kk1vc88UBkajhOlq0MqW2ty00xLNIwCCEGT-ORWRYePA> <xmx:Bvb7XvzVRziVpLp6iZmCM_0lWS_u-usfn5sf67Ma6BQYr__faIbyoQ>
Received: by mailuser.nyi.internal (Postfix, from userid 501) id 66B49E00CE; Tue, 30 Jun 2020 22:33:42 -0400 (EDT)
X-Mailer: MessagingEngine.com Webmail Interface
User-Agent: Cyrus-JMAP/3.3.0-dev0-576-gfe2cd66-fm-20200629.001-gfe2cd668
Mime-Version: 1.0
Message-Id: <7b0c8566-46eb-490a-8b93-3fe7eb3e9ee4@www.fastmail.com>
Date: Wed, 01 Jul 2020 12:33:24 +1000
From: "Martin Thomson" <mt@lowentropy.net>
To: quic@ietf.org
Subject: Versions numbers in invariants
Content-Type: text/plain
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/Zqm5cQL0Qeaa1YjYL1o7ZfQPhig>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 01 Jul 2020 02:33:45 -0000

In my review of QUIC Invariants, I think that I found a promise we might not want to keep:

> QUIC versions are identified with a 32-bit integer, encoded in network byte order.

Implicit in this is that a packet containing version X is necessarily QUIC version X.  That's already not-true in some proposals.

I think that this would be better defined as a field for use by endpoints:

> The Version field contains a 32-bit integer, encoded in network byte order.
> This value can be used by endpoints to identify a QUIC Version.

This is an almost-cosmetic change, but it is a substantive change.  The intent is to wind back this promise and allow for the sorts of mappings contemplated by the version aliasing proposal.

https://github.com/quicwg/base-drafts/issues/3808 is the issue, with some more details.
https://github.com/quicwg/base-drafts/issues/3785 contains the proposed change.


From nobody Tue Jun 30 21:12:38 2020
Return-Path: <mikkelfj@gmail.com>
X-Original-To: quic@ietfa.amsl.com
Delivered-To: quic@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C7EC63A0BA7 for <quic@ietfa.amsl.com>; Tue, 30 Jun 2020 21:12:37 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.096
X-Spam-Level: 
X-Spam-Status: No, score=-2.096 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, UNPARSEABLE_RELAY=0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id sKzUKMsHJLHJ for <quic@ietfa.amsl.com>; Tue, 30 Jun 2020 21:12:35 -0700 (PDT)
Received: from mail-yb1-xb35.google.com (mail-yb1-xb35.google.com [IPv6:2607:f8b0:4864:20::b35]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id C1F683A0B6A for <quic@ietf.org>; Tue, 30 Jun 2020 21:12:35 -0700 (PDT)
Received: by mail-yb1-xb35.google.com with SMTP id y17so9175131ybm.12 for <quic@ietf.org>; Tue, 30 Jun 2020 21:12:35 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;  h=from:in-reply-to:references:mime-version:date:message-id:subject:to :cc; bh=AemDyWHA4LLEq5eQs2ue5tQmpVkgsoMZRFSpZhIuScQ=; b=f+hUJhsv6FbLfAC3hWvFZRhTR8oZuCV2V635ZsRag7iJoezD8sJBwEG2gkS1xyzHnm b58ipq2UJwb+6HqVxIe+4nAckBhBcjQRRm61sS2EGhplo8Qf2JIfXcz+8ZxjwuV3bCx2 yq7TbnHKJk/0KeViU+LxAmXr3PoJLeDAXZL8EOgt9pqJBOiDaWdLmCLuG5xQBJE4nUv/ ZAmnmAmggmUJavKLOVQivYbWzpRA+E6/sMq5qGnQn3RTcJNpst3jbldEFxMzKVtv/CvN 4MLe9rda6svCa7qxztenY1ZivgJwC+tLs8z0l5a5Ce/ZngE+fllCeEhbJDhiKNu6Kw3l Y1wg==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:in-reply-to:references:mime-version:date :message-id:subject:to:cc; bh=AemDyWHA4LLEq5eQs2ue5tQmpVkgsoMZRFSpZhIuScQ=; b=D7UslLyX3a3s27UbMAhUJehuy7hoJ/9tVuJS0DFyESsTRGmG1xbpsKrItcCEyHkVLH KL0t4Utm6+3U/T8uS6fxREJWS1Uhz17XSVOPsPlKEKhtIFNchLoYpLwOtfWlm0ttQBAb aj8Vz0OGTIKbnuLc5UQsD+Srb90ZKPHONZDBxOtAzA+ui9bH4ILBAtSnzt+AJRqNOP+f Tr5HmzAe53waIVnE2A4inkF8vxL9sBljaLYoKONix5NSSQznGvVTq8bHaf+G+QkQwIPx il+SIdcEQieazq/JEuf6jLdI9lQm34NidXXkLwCulFcPIkV1fLFMb75C0qpO2hLLXP3V +Bug==
X-Gm-Message-State: AOAM532MNfAWqIiP5KuZf2RnRSOegAbz7hjYcVxK3kd70WiDwWKItu30 4tZ+qfMfgj7xEGUe1KyPh7rWaUQnsl7HjWOA+P0=
X-Google-Smtp-Source: ABdhPJxt66kEvfdO3+qZupHtJngou5y4Bl/wvUkVD1HjkOvmLusfeGARfPIT3sZ+5eSmIScedkh5+R/y6ZdAsR3cTJQ=
X-Received: by 2002:a25:fd5:: with SMTP id 204mr38484053ybp.294.1593576754816;  Tue, 30 Jun 2020 21:12:34 -0700 (PDT)
Received: from 1058052472880 named unknown by gmailapi.google.com with HTTPREST; Tue, 30 Jun 2020 21:12:34 -0700
From: =?UTF-8?Q?Mikkel_Fahn=C3=B8e_J=C3=B8rgensen?= <mikkelfj@gmail.com>
In-Reply-To: <E86FC27B-E38E-4842-BD4F-A316ADD51FE9@eggert.org>
References: <f97dbbe6-e719-4276-bd19-da6589bb2d4e@www.fastmail.com> <E86FC27B-E38E-4842-BD4F-A316ADD51FE9@eggert.org>
MIME-Version: 1.0
Date: Tue, 30 Jun 2020 21:12:34 -0700
Message-ID: <CAN1APdejsLess_eLQPnaUOWhfiZXFiGG07poGYvtQE=9xq1bTw@mail.gmail.com>
Subject: Re: Padding in Initial Packets
To: Lars Eggert <lars@eggert.org>, Martin Thomson <mt@lowentropy.net>
Cc: quic@ietf.org
Content-Type: multipart/alternative; boundary="00000000000064aecb05a9598078"
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic/rm5Th5g_PTZRuwWkiiSt25aOIqM>
X-BeenThere: quic@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Main mailing list of the IETF QUIC working group <quic.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/quic>, <mailto:quic-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/quic/>
List-Post: <mailto:quic@ietf.org>
List-Help: <mailto:quic-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/quic>, <mailto:quic-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 01 Jul 2020 04:12:38 -0000

--00000000000064aecb05a9598078
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Maybe an an amplification attack is possible by sending all zeroes, if it
happens over a compressed link as opposed to sending random data. Maybe a
sat uplink of sorts.

This could (mostly) be avoided be having all initial data inside a QUIC
packet, although a really aggressive compressor could figure this out for
the initial packet.

Mikkel

On 1 July 2020 at 04.17.24, Lars Eggert (lars@eggert.org) wrote:

Quant similarly pads by coalescing random data (with the first byte
massaged to be of short-header type).

--=20
Sent from a mobile device; please excuse typos.

> On Jul 1, 2020, at 04:19, Martin Thomson <mt@lowentropy.net> wrote:
>
> =EF=BB=BFHi Tim,
>
> That's a little curveball that Firefox throws. Other implementations pad
with PADDING frames inside Initial packets. Our first implementation found
that inconvenient and so we padded outside the packet. The resulting
"packets" are clearly not valid, but neither server nor spec cannot require
that they be legitimate. QUIC doesn't authenticate the composition of QUIC
packets into UDP datagrams. So the overall result is valid, even if the
individual bits are not.
>
> The number of bugs this behaviour has exposed is so great that we
ultimately decided to keep it. Fixing it would make our code more
complicated anyway.
>
> We might change the value we use for padding from 0 to something else at
some point, just to keep things fresh.
>
> Cheers,
> Martin
>
>> On Tue, Jun 30, 2020, at 20:15, tim.jebb@bt.com wrote:
>>
>> Hi All,
>>
>>
>> I=E2=80=99m fairly new to QUIC so apologies if the questions in here are=
 dumb.
>>
>>
>> I=E2=80=99ve been going through a QUIC pcap, which I captured from a Fir=
efox
>> implementation of IETF draft 27 QUIC, and am a little confused about
>> the use of padding in the first Initial packet. According to Wireshark,
>> the UDP datagram contains a QUIC packet of type initial, length 272,
>> the payload of which is a Client Hello.
>>
>>
>> The remainder of the UDP datagram is 1055 zero bytes.. According to
>> Wireshark these zero bytes constitute a QUIC IETF Packet.
>>
>>
>> However there doesn=E2=80=99t seem to be a valid QUIC packet header for =
these
>> bytes. AIUI according to the QUIC spec padding bytes are frames within
>> a packet, so these zero bytes seem to be just bytes in the UDP datagram
>> outside of the QUIC payload. Which would be fine as far as I can see,
>> but:
>>
>>
>> Why does Wireshark say it=E2=80=99s QUIC IETF?
>>
>> Is there anything in any of the QUIC draft RFCs mandating this
>> behaviour, if so where?
>>
>> Why is it being done =E2=80=93 can=E2=80=99t see the point?
>>
>> If padding is to be added surely it should be as part of a QUIC packet
>> and encrypted =E2=80=93 here there=E2=80=99s just a load of plaintext ze=
ro bytes.
>>
>>
>>
>> Here is Wireshark=E2=80=99s analysis:
>>
>>
>> =E2=80=9CQUIC IETF
>>
>> QUIC Connection information
>>
>> [Packet Length: 302]
>>
>> 1... .... =3D Header Form: Long Header (1)
>>
>> .1.. .... =3D Fixed Bit: True
>>
>> ..00 .... =3D Packet Type: Initial (0)
>>
>> .... 00.. =3D Reserved: 0
>>
>> .... ..10 =3D Packet Number Length: 3 bytes (2)
>>
>> Version: draft-27 (0xff00001b)
>>
>> Destination Connection ID Length: 17
>>
>> Destination Connection ID: 76f81a5588b2f25c5ab6dbd461a8d749e1
>>
>> Source Connection ID Length: 3
>>
>> Source Connection ID: c5ef02
>>
>> Token Length: 0
>>
>> Length: 272
>>
>> Packet Number: 0
>>
>> Payload: 04f535177fcbcfd5f06d24ffde66edf10df79205666b1851=E2=80=A6
>>
>> TLSv1.3 Record Layer: Handshake Protocol: Client Hello
>>
>> Frame Type: CRYPTO (0x0000000000000006)
>>
>> Offset: 0
>>
>> Length: 249
>>
>> Crypto Data
>>
>> Handshake Protocol: Client Hello
>>
>> Handshake Type: Client Hello (1)
>>
>> Length: 245
>>
>> Version: TLS 1.2 (0x0303)
>>
>> Random: b1b76126ddab3c639cdfaa7d92d6cabe2ea9c1a19ea86b8b=E2=80=A6
>>
>> Session ID Length: 0
>>
>> Cipher Suites Length: 4
>>
>> Cipher Suites (2 suites)
>>
>> Compression Methods Length: 1
>>
>> Compression Methods (1 method)
>>
>> Extensions Length: 200
>>
>> Extension: server_name (len=3D15)
>>
>> Extension: extended_master_secret (len=3D0)
>>
>> Extension: renegotiation_info (len=3D1)
>>
>> Extension: supported_groups (len=3D20)
>>
>> Extension: application_layer_protocol_negotiation (len=3D8)
>>
>> Extension: status_request (len=3D5)
>>
>> Extension: key_share (len=3D38)
>>
>> Extension: supported_versions (len=3D3)
>>
>> Extension: signature_algorithms (len=3D16)
>>
>> Extension: psk_key_exchange_modes (len=3D2)
>>
>> Extension: record_size_limit (len=3D2)
>>
>> Extension: quic_transports_parameters (len=3D42)
>>
>> QUIC IETF
>>
>> [Packet Length: 1055]
>>
>> Remaining Payload: 000000000000000000000000000000000000000000000000=E2=
=80=A6=E2=80=9D
>>
>>
>> Any comments or explanations?
>>
>>
>> Tim
>>
>

--00000000000064aecb05a9598078
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<html><head><style>body{font-family:Helvetica,Arial;font-size:13px}</style>=
</head><body><div style=3D"font-family:Helvetica,Arial;font-size:13px">Mayb=
e an an amplification attack is possible by sending all zeroes, if it happe=
ns over a compressed link as opposed to sending random data. Maybe a sat up=
link of sorts.</div><div style=3D"font-family:Helvetica,Arial;font-size:13p=
x"><br></div><div style=3D"font-family:Helvetica,Arial;font-size:13px">This=
 could (mostly) be avoided be having all initial data inside a QUIC packet,=
 although a really aggressive compressor could figure this out for the init=
ial packet.</div><div style=3D"font-family:Helvetica,Arial;font-size:13px">=
<br></div> <div class=3D"gmail_signature"><div style=3D"font-size:13px">Mik=
kel</div></div> <br><p class=3D"airmail_on">On 1 July 2020 at 04.17.24, Lar=
s Eggert (<a href=3D"mailto:lars@eggert.org">lars@eggert.org</a>) wrote:</p=
> <blockquote type=3D"cite" class=3D"clean_bq"><span><div><div></div><div>Q=
uant similarly pads by coalescing random data (with the first byte massaged=
 to be of short-header type).
<br>
<br>-- =20
<br>Sent from a mobile device; please excuse typos.
<br>
<br>&gt; On Jul 1, 2020, at 04:19, Martin Thomson &lt;<a href=3D"mailto:mt@=
lowentropy.net">mt@lowentropy.net</a>&gt; wrote:
<br>&gt; =20
<br>&gt; =EF=BB=BFHi Tim,
<br>&gt; =20
<br>&gt; That&#39;s a little curveball that Firefox throws.  Other implemen=
tations pad with PADDING frames inside Initial packets.  Our first implemen=
tation found that inconvenient and so we padded outside the packet.  The re=
sulting &quot;packets&quot; are clearly not valid, but neither server nor s=
pec cannot require that they be legitimate. QUIC doesn&#39;t authenticate t=
he composition of QUIC packets into UDP datagrams.  So the overall result i=
s valid, even if the individual bits are not.
<br>&gt; =20
<br>&gt; The number of bugs this behaviour has exposed is so great that we =
ultimately decided to keep it.  Fixing it would make our code more complica=
ted anyway.
<br>&gt; =20
<br>&gt; We might change the value we use for padding from 0 to something e=
lse at some point, just to keep things fresh.
<br>&gt; =20
<br>&gt; Cheers,
<br>&gt; Martin
<br>&gt; =20
<br>&gt;&gt; On Tue, Jun 30, 2020, at 20:15, <a href=3D"mailto:tim.jebb@bt.=
com">tim.jebb@bt.com</a> wrote:
<br>&gt;&gt; =20
<br>&gt;&gt; Hi All,
<br>&gt;&gt; =20
<br>&gt;&gt; =20
<br>&gt;&gt; I=E2=80=99m fairly new to QUIC so apologies if the questions i=
n here are dumb.
<br>&gt;&gt; =20
<br>&gt;&gt; =20
<br>&gt;&gt; I=E2=80=99ve been going through a QUIC pcap, which I captured =
from a Firefox =20
<br>&gt;&gt; implementation of IETF draft 27 QUIC, and am a little confused=
 about =20
<br>&gt;&gt; the use of padding in the first Initial packet. According to W=
ireshark, =20
<br>&gt;&gt; the UDP datagram contains a QUIC packet of type initial, lengt=
h 272, =20
<br>&gt;&gt; the payload of which is a Client Hello. =20
<br>&gt;&gt; =20
<br>&gt;&gt; =20
<br>&gt;&gt; The remainder of the UDP datagram is 1055 zero bytes.. Accordi=
ng to =20
<br>&gt;&gt; Wireshark these zero bytes constitute a QUIC IETF Packet.
<br>&gt;&gt; =20
<br>&gt;&gt; =20
<br>&gt;&gt; However there doesn=E2=80=99t seem to be a valid QUIC packet h=
eader for these =20
<br>&gt;&gt; bytes. AIUI according to the QUIC spec padding bytes are frame=
s within =20
<br>&gt;&gt; a packet, so these zero bytes seem to be just bytes in the UDP=
 datagram =20
<br>&gt;&gt; outside of the QUIC payload. Which would be fine as far as I c=
an see, =20
<br>&gt;&gt; but:
<br>&gt;&gt; =20
<br>&gt;&gt; =20
<br>&gt;&gt; Why does Wireshark say it=E2=80=99s QUIC IETF?
<br>&gt;&gt; =20
<br>&gt;&gt; Is there anything in any of the QUIC draft RFCs mandating this=
 =20
<br>&gt;&gt; behaviour, if so where?
<br>&gt;&gt; =20
<br>&gt;&gt; Why is it being done =E2=80=93 can=E2=80=99t see the point?
<br>&gt;&gt; =20
<br>&gt;&gt; If padding is to be added surely it should be as part of a QUI=
C packet =20
<br>&gt;&gt; and encrypted =E2=80=93 here there=E2=80=99s just a load of pl=
aintext zero bytes.
<br>&gt;&gt; =20
<br>&gt;&gt; =20
<br>&gt;&gt; =20
<br>&gt;&gt; Here is Wireshark=E2=80=99s analysis:
<br>&gt;&gt; =20
<br>&gt;&gt; =20
<br>&gt;&gt; =E2=80=9CQUIC IETF
<br>&gt;&gt; =20
<br>&gt;&gt; QUIC Connection information
<br>&gt;&gt; =20
<br>&gt;&gt; [Packet Length: 302]
<br>&gt;&gt; =20
<br>&gt;&gt; 1... .... =3D Header Form: Long Header (1)
<br>&gt;&gt; =20
<br>&gt;&gt; .1.. .... =3D Fixed Bit: True
<br>&gt;&gt; =20
<br>&gt;&gt; ..00 .... =3D Packet Type: Initial (0)
<br>&gt;&gt; =20
<br>&gt;&gt; .... 00.. =3D Reserved: 0
<br>&gt;&gt; =20
<br>&gt;&gt; .... ..10 =3D Packet Number Length: 3 bytes (2)
<br>&gt;&gt; =20
<br>&gt;&gt; Version: draft-27 (0xff00001b)
<br>&gt;&gt; =20
<br>&gt;&gt; Destination Connection ID Length: 17
<br>&gt;&gt; =20
<br>&gt;&gt; Destination Connection ID: 76f81a5588b2f25c5ab6dbd461a8d749e1
<br>&gt;&gt; =20
<br>&gt;&gt; Source Connection ID Length: 3
<br>&gt;&gt; =20
<br>&gt;&gt; Source Connection ID: c5ef02
<br>&gt;&gt; =20
<br>&gt;&gt; Token Length: 0
<br>&gt;&gt; =20
<br>&gt;&gt; Length: 272
<br>&gt;&gt; =20
<br>&gt;&gt; Packet Number: 0
<br>&gt;&gt; =20
<br>&gt;&gt; Payload: 04f535177fcbcfd5f06d24ffde66edf10df79205666b1851=E2=
=80=A6
<br>&gt;&gt; =20
<br>&gt;&gt; TLSv1.3 Record Layer: Handshake Protocol: Client Hello
<br>&gt;&gt; =20
<br>&gt;&gt; Frame Type: CRYPTO (0x0000000000000006)
<br>&gt;&gt; =20
<br>&gt;&gt; Offset: 0
<br>&gt;&gt; =20
<br>&gt;&gt; Length: 249
<br>&gt;&gt; =20
<br>&gt;&gt; Crypto Data
<br>&gt;&gt; =20
<br>&gt;&gt; Handshake Protocol: Client Hello
<br>&gt;&gt; =20
<br>&gt;&gt; Handshake Type: Client Hello (1)
<br>&gt;&gt; =20
<br>&gt;&gt; Length: 245
<br>&gt;&gt; =20
<br>&gt;&gt; Version: TLS 1.2 (0x0303)
<br>&gt;&gt; =20
<br>&gt;&gt; Random: b1b76126ddab3c639cdfaa7d92d6cabe2ea9c1a19ea86b8b=E2=80=
=A6
<br>&gt;&gt; =20
<br>&gt;&gt; Session ID Length: 0
<br>&gt;&gt; =20
<br>&gt;&gt; Cipher Suites Length: 4
<br>&gt;&gt; =20
<br>&gt;&gt; Cipher Suites (2 suites)
<br>&gt;&gt; =20
<br>&gt;&gt; Compression Methods Length: 1
<br>&gt;&gt; =20
<br>&gt;&gt; Compression Methods (1 method)
<br>&gt;&gt; =20
<br>&gt;&gt; Extensions Length: 200
<br>&gt;&gt; =20
<br>&gt;&gt; Extension: server_name (len=3D15)
<br>&gt;&gt; =20
<br>&gt;&gt; Extension: extended_master_secret (len=3D0)
<br>&gt;&gt; =20
<br>&gt;&gt; Extension: renegotiation_info (len=3D1)
<br>&gt;&gt; =20
<br>&gt;&gt; Extension: supported_groups (len=3D20)
<br>&gt;&gt; =20
<br>&gt;&gt; Extension: application_layer_protocol_negotiation (len=3D8)
<br>&gt;&gt; =20
<br>&gt;&gt; Extension: status_request (len=3D5)
<br>&gt;&gt; =20
<br>&gt;&gt; Extension: key_share (len=3D38)
<br>&gt;&gt; =20
<br>&gt;&gt; Extension: supported_versions (len=3D3)
<br>&gt;&gt; =20
<br>&gt;&gt; Extension: signature_algorithms (len=3D16)
<br>&gt;&gt; =20
<br>&gt;&gt; Extension: psk_key_exchange_modes (len=3D2)
<br>&gt;&gt; =20
<br>&gt;&gt; Extension: record_size_limit (len=3D2)
<br>&gt;&gt; =20
<br>&gt;&gt; Extension: quic_transports_parameters (len=3D42)
<br>&gt;&gt; =20
<br>&gt;&gt; QUIC IETF
<br>&gt;&gt; =20
<br>&gt;&gt; [Packet Length: 1055]
<br>&gt;&gt; =20
<br>&gt;&gt; Remaining Payload: 0000000000000000000000000000000000000000000=
00000=E2=80=A6=E2=80=9D
<br>&gt;&gt; =20
<br>&gt;&gt; =20
<br>&gt;&gt; Any comments or explanations?
<br>&gt;&gt; =20
<br>&gt;&gt; =20
<br>&gt;&gt; Tim
<br>&gt;&gt; =20
<br>&gt; =20
<br>
<br></div></div></span></blockquote></body></html>

--00000000000064aecb05a9598078--

