Return-Path: X-Original-To: ietfarch-httpbisa-archive-bis2Juki@ietfa.amsl.com Delivered-To: ietfarch-httpbisa-archive-bis2Juki@ietfa.amsl.com Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 9426812DF24 for ; Thu, 18 Aug 2016 07:11:43 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -8.167 X-Spam-Level: X-Spam-Status: No, score=-8.167 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, HEADER_FROM_DIFFERENT_DOMAINS=0.001, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_HI=-5, RCVD_IN_MSPIKE_H3=-0.01, RCVD_IN_MSPIKE_WL=-0.01, RP_MATCHES_RCVD=-1.247, SPF_HELO_PASS=-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=sendgrid.me 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 y2UCK-VH2Eaw for ; Thu, 18 Aug 2016 07:11:41 -0700 (PDT) Received: from frink.w3.org (frink.w3.org [128.30.52.56]) (using TLSv1.2 with cipher DHE-RSA-AES128-SHA (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 2813912DF1E for ; Thu, 18 Aug 2016 07:11:41 -0700 (PDT) Received: from lists by frink.w3.org with local (Exim 4.80) (envelope-from ) id 1baNyX-0007hA-N0 for ietf-http-wg-dist@listhub.w3.org; Thu, 18 Aug 2016 14:07:17 +0000 Resent-Date: Thu, 18 Aug 2016 14:07:17 +0000 Resent-Message-Id: Received: from maggie.w3.org ([128.30.52.39]) by frink.w3.org with esmtps (TLS1.2:DHE_RSA_AES_128_CBC_SHA1:128) (Exim 4.80) (envelope-from ) id 1baNyQ-0007gP-Cd for ietf-http-wg@listhub.w3.org; Thu, 18 Aug 2016 14:07:10 +0000 Received: from o1.30e.fshared.sendgrid.net ([167.89.55.41]) by maggie.w3.org with esmtps (TLS1.2:DHE_RSA_AES_128_CBC_SHA256:128) (Exim 4.80) (envelope-from ) id 1baNyJ-00010f-Py for ietf-http-wg@w3.org; Thu, 18 Aug 2016 14:07:06 +0000 DKIM-Signature: v=1; a=rsa-sha1; c=relaxed; d=sendgrid.me; h=mime-version:in-reply-to:references:from:subject:to:cc:content-type; s=smtpapi; bh=Fc6kSpcQVI7rgxqjcu2DydTKvTA=; b=GxR0Q4HtstpWYP+tGl x8gY4innuNw0X4DBi7WuoDApFnC45886+6lzsAKNZhATq8A5ilpnn38g3QDw1tkB uphBRdcikW0OEBL+32TucyI2b7t03lNKa1e/AR73p5SByky9NAQAwhqjFi3KU80+ zxFcmwNebaIq4AO7fYgkkGrGY= Received: by filter0936p1mdw1.sendgrid.net with SMTP id filter0936p1mdw1.9172.57B5C0E6A4 2016-08-18 14:06:30.969551733 +0000 UTC Received: from mail-ua0-f170.google.com (mail-ua0-f170.google.com [209.85.217.170]) by ismtpd0003p1iad1.sendgrid.net (SG) with ESMTP id U9PMDPO5TXy45QsyhaWeyQ for ; Thu, 18 Aug 2016 14:06:30.926 +0000 (UTC) Received: by mail-ua0-f170.google.com with SMTP id n59so29592658uan.2 for ; Thu, 18 Aug 2016 07:06:30 -0700 (PDT) X-Gm-Message-State: AEkoouuNYhoVdQ5jCXEbdKPQuH+iwZNBX9BvY0VGZH+hG/JhIpIRkv4sTISgXKnRwx9jBpJ8HzQI4rdONmgw2A== X-Received: by 10.159.53.106 with SMTP id o97mr1193549uao.40.1471529190705; Thu, 18 Aug 2016 07:06:30 -0700 (PDT) MIME-Version: 1.0 Received: by 10.176.81.242 with HTTP; Thu, 18 Aug 2016 07:06:29 -0700 (PDT) In-Reply-To: References: <3EE9A02C-794A-4147-A108-914AB19F2800@lukasa.co.uk> <56290C49.6040301@crf.canon.fr> <20151023161519.GA26338@LK-Perkele-V2.elisa-laajakaista.fi> From: Patrick McManus Date: Thu, 18 Aug 2016 10:06:29 -0400 X-Gmail-Original-Message-ID: Message-ID: To: Tatsuhiro Tsujikawa Cc: =?UTF-8?Q?Bence_B=C3=A9ky?= , "ietf-http-wg@w3.org" , =?UTF-8?Q?Herv=C3=A9_Ruellan?= , Cory Benfield , Ilari Liusvaara Content-Type: multipart/alternative; boundary=94eb2c03ed54b052b1053a5916dd X-SG-EID: YLWet4rakcOTMHWvPPwWbcsiUJbN1FCn0PHYd/Uujh70YbPEZROubAew6dxleUfW/VWmUlg4jTYntT ag8YAttVkdBMCw0fDBQVRZWJWPSyXI5KvvicTamdakBTGi42ZUqTG55wPLvS2bcJsr/OzCwCrpjacZ KQl0nhzrAnJRyAph6URTq9xH3RbKUVvdgeaZLEgqytRxjtKIo3zrE6fFraywUpQpQsLgfrlPNKT81J s= Received-SPF: pass client-ip=167.89.55.41; envelope-from=bounces+1568871-208f-ietf-http-wg=w3.org@sendgrid.net; helo=o1.30e.fshared.sendgrid.net X-W3C-Hub-Spam-Status: No, score=-5.7 X-W3C-Hub-Spam-Report: AWL=-0.635, BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_NONE=-0.0001, RP_MATCHES_RCVD=-0.548, SPF_PASS=-0.001, URIBL_GREY=0.424, W3C_AA=-1, W3C_IRA=-1, W3C_WL=-1 X-W3C-Scan-Sig: maggie.w3.org 1baNyJ-00010f-Py 3051909874e47008f4c09fba31054f13 X-Original-To: ietf-http-wg@w3.org Subject: Re: Clarification of dynamic table size change Archived-At: Resent-From: ietf-http-wg@w3.org X-Mailing-List: archive/latest/32315 X-Loop: ietf-http-wg@w3.org Resent-Sender: ietf-http-wg-request@w3.org Precedence: list List-Id: List-Help: List-Post: List-Unsubscribe: --94eb2c03ed54b052b1053a5916dd Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Tatsuhiro's description matches my understanding. -Patrick On Thu, Aug 18, 2016 at 9:50 AM, Tatsuhiro Tsujikawa wrote: > =E2=80=8BHi,=E2=80=8B > > On Thu, Aug 18, 2016 at 10:31 PM, Bence B=C3=A9ky wrot= e: > >> Hi all, >> >> I'm sorry to revive this old thread, but there is one more case that I >> would >> like to request clarification for. I was looking at both RFC 7540 and >> 7541, but >> could not find a definitive answer to the following question: What is t= he >> initial maximum size of the dynamic table if there was a >> SETTINGS_HEADER_TABLE_SIZE value in the initial SETTINGS frame (the one >> part of >> the connection preface)? >> >> For example, suppose that the decoder sends a SETTINGS_HEADER_TABLE with >> value >> 64 * 1024 in the initial SETTINGS frame. Do we think about the HPACK >> context to >> be created after the connection preface is sent, with a maximum dynamic >> table >> size of the current SETTINGS_HEADER_TABLE_SIZE value of 64 kB? Or do we >> think >> about the HPACK context to be created before the connection preface is >> sent, >> with a default maximum dynamic table size of 4 kB? Note that there is no >> synchronization issue even in the former case: if the decoder only evicts >> dynamic table entries above 64 kB from the very beginning, there is no >> harm in >> the encoder not starting to reference entries above 4 kB until it >> processes the >> decoder's initial SETTINGS frame. >> >> Suppose that the encoder does not emit a "dynamic table size update" HPA= CK >> instruction after this. The consensus on this e-mail thread seems to be >> that >> this is acceptable as long as the encoder means "no change" to the maxim= um >> dynamic table size. It is, however, important that the encoder and the >> decoder >> are in agreement about the initial maximum dynamic table size, relative >> to which >> the encoder means "no change". For example, if the decoder is under the >> impression that the maximum dynamic table size is 4 kB, while the encoder >> takes >> it to be 64 kB, then the decoder will signal a CONNECTION_ERROR as soon >> as the >> encoder references an entry above 4 kB. If, on the other hand, the >> encoder >> thinks it's 4 kB and never references entries above that, then the >> decoder would >> waste memory if it kept 64 kB worth of entries. >> >> Given that a decoder can send a SETTINGS_HEADER_TABLE_SIZE with a value >> lower >> than the default, and the encoder can start compressing headers before >> receiving >> the initial SETTINGS frame, it seems necessary to me to understand the >> initial >> maximum dynamic table size to be 4 kB, and to require the decoder to >> store this >> much entries until it receives the dynamic table size update HPACK >> instruction >> from the encoder. Otherwise a COMPRESSION_ERROR arises due to the >> synchronization issue even if the peers agree that the initial size is >> the new >> (lower) value. Unless, of course, we want to formulate different >> requirements >> depending on whether the SETTINGS_HEADER_TABLE_SIZE value is greater than >> or >> less than the default. >> >> If I implement a decoder in this spirit, that is, one that sends a >> SETTINGS_HEADER_TABLE_SIZE of 64 kB in the initial SETTINGS frame, but >> does not >> allow more than default memory for the dynamic table until it receives a >> dynamic >> table size update from the encoder, would it be incompatible with >> anybody's >> current implementation? >> >> > According to this thread, =E2=80=8BI'm under impression =E2=80=8Bthat thi= s is OK, and > until you get dynamic table size update, default 4KiB dynamic table limit > still applies. > > As for initial value of dynamic table size, I think it is 4KiB regardless > of SETTINGS. We create HTTP/2 session before doing any parameter > modification, including header table size change. At this moment, table > size if 4KiB, RFC default. After that, decoder send > SETTINGS_HEADER_TABLE_SIZE with whatever value they want. Then after > SETTINGS ACK, and HPACK table size update, dynamic table size is finally > synchronized, and changed to the value encoder sent in HPACK table size > update (as long as it is equal or smaller than decoder sent in SETTINGS). > > Best regards, > Tatsuhiro Tsujikawa > > > > >> Best regards, >> >> Bence B=C3=A9ky >> >> >> On Fri, Oct 23, 2015 at 12:15 PM, Ilari Liusvaara < >> ilariliusvaara@welho.com> wrote: >> >>> On Sat, Oct 24, 2015 at 12:45:49AM +0900, Tatsuhiro Tsujikawa wrote: >>> > Hi, >>> > >>> > On Fri, Oct 23, 2015 at 1:18 AM, Herv=C3=A9 Ruellan < >>> herve.ruellan@crf.canon.fr> >>> > wrote: >>> > >>> > > I agree that the wording is ambiguous here. >>> > > >>> > > However, my reading is the same a Cory's: you don't have to send a >>> dynamic >>> > > table update if the *actual* value is not changed. >>> > > >>> > > >>> > =E2=80=8BI also found the discussion in this ML indicating you are ri= ght. >>> Thank >>> > you for clarification. >>> > I have to ask one more question: what is *actual* value? Is it the >>> table >>> > size both peer agreed before reading SETTINGS, or the value in >>> > SETTINGS_HEADER_TABLE_SIZE decoder sent? >>> > >>> > I think this is a good item to add in FAQ section.. >>> >>> The way negotiation works: >>> - Decoder side sets the upper bound via SETTINGS_HEADER_TABLE_SIZE. >>> - Encoder side sets the actual size via dynamic table updates (inside >>> HPACK bitstream) within limits set by decoder. >>> - If between headers decoder reduces the limit below size signaled by >>> encoder, the encoder must first reduce the table size to the minimum >>> it was between the frames or less (it can then increase it up to >>> current limit). >>> >>> As example of the last point: >>> [4k dynamic table size in use] >>> --> HEADERS >>> <-- SETTINGS(SETTINGS_HEADER_TABLE_SIZE=3D4k) >>> <-- SETTINGS(SETTINGS_HEADER_TABLE_SIZE=3D2k) >>> <-- SETTINGS(SETTINGS_HEADER_TABLE_SIZE=3D4k) >>> <-- SETTINGS(SETTINGS_HEADER_TABLE_SIZE=3D8k) >>> <-- SETTINGS(SETTINGS_HEADER_TABLE_SIZE=3D6k) >>> --> HEADERS >>> >>> The second HEADERS must first reduce the dynamic table to at most >>> 2k. It can then increase dynamic table size to up to 6k. >>> >>> >>> -Ilari >>> >>> >> > --94eb2c03ed54b052b1053a5916dd Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
Tatsuhiro's description matches my understanding.=

-Patrick


On Thu, Aug 18, 2016 at 9:50 AM, Tatsuhiro Tsujikawa <tatsuhiro.t@gmail.com> wrote:
=E2=80= =8BHi,=E2=80=8B

= On Thu, Aug 18, 2016 at 10:31 PM, Bence B=C3=A9ky <bnc@chromium.org>= wrote:
Hi a= ll,

I'm sorry to revive this old thread, but t= here is one more case that I would
like to request clarification = for.=C2=A0 I was looking at both RFC 7540 and 7541, but
could not= find a definitive answer to the following question: =C2=A0What is the
initial maximum size of the dynamic table if there was a
SE= TTINGS_HEADER_TABLE_SIZE value in the initial SETTINGS frame (the one part = of
the connection preface)?

For example,= suppose that the decoder sends a SETTINGS_HEADER_TABLE with value
64 * 1024 in the initial SETTINGS frame.=C2=A0 Do we think about the HPAC= K context to
be created after the connection preface is sent, wit= h a maximum dynamic table
size of the current SETTINGS_HEADER_TAB= LE_SIZE value of 64 kB?=C2=A0 Or do we think
about the HPACK cont= ext to be created before the connection preface is sent,
with a d= efault maximum dynamic table size of 4 kB?=C2=A0 Note that there is no
synchronization issue even in the former case: if the decoder only ev= icts
dynamic table entries above 64 kB from the very beginning, t= here is no harm in
the encoder not starting to reference entries = above 4 kB until it processes the
decoder's initial SETTINGS = frame.

Suppose that the encoder does not emit a &q= uot;dynamic table size update" HPACK
instruction after this.= =C2=A0 The consensus on this e-mail thread seems to be that
this = is acceptable as long as the encoder means "no change" to the max= imum
dynamic table size.=C2=A0 It is, however, important that the= encoder and the decoder
are in agreement about the initial maxim= um dynamic table size, relative to which
the encoder means "= no change".=C2=A0 For example, if the decoder is under the
i= mpression that the maximum dynamic table size is 4 kB, while the encoder ta= kes
it to be 64 kB, then the decoder will signal a CONNECTION_ERR= OR as soon as the
encoder references an entry above 4 kB.=C2=A0 I= f, on the other hand, the encoder
thinks it's 4 kB and never = references entries above that, then the decoder would
waste memor= y if it kept 64 kB worth of entries.

Given that a = decoder can send a SETTINGS_HEADER_TABLE_SIZE with a value lower
= than the default, and the encoder can start compressing headers before rece= iving
the initial SETTINGS frame, it seems necessary to me to und= erstand the initial
maximum dynamic table size to be 4 kB, and to= require the decoder to store this
much entries until it receives= the dynamic table size update HPACK instruction
from the encoder= .=C2=A0 Otherwise a COMPRESSION_ERROR arises due to the
synchroni= zation issue even if the peers agree that the initial size is the new
=
(lower) value.=C2=A0 Unless, of course, =C2=A0we want to formulate dif= ferent requirements
depending on whether the SETTINGS_HEADER_TABL= E_SIZE value is greater than or
less than the default.
=
If I implement a decoder in this spirit, that is, one that s= ends a
SETTINGS_HEADER_TABLE_SIZE of 64 kB in the initial SETTING= S frame, but does not
allow more than default memory for the dyna= mic table until it receives a dynamic
table size update from the = encoder, would it be incompatible with anybody's
current impl= ementation?


According to this thread, =E2=80=8BI'm under impression= =E2=80=8Bthat this is OK, and until you get dynamic table size update, def= ault 4KiB dynamic table limit still applies.

As for initial value of dynamic table size, I think it is 4KiB regardless= of SETTINGS.=C2=A0 We create HTTP/2 session before doing any parameter mod= ification, including header table size change.=C2=A0 At this moment, table = size if 4KiB, RFC default.=C2=A0 After that, decoder send SETTINGS_HEADER_T= ABLE_SIZE with whatever value they want.=C2=A0 Then after SETTINGS ACK, and= HPACK table size update, dynamic table size is finally synchronized, and c= hanged to the value encoder sent in HPACK table size update (as long as it = is equal or smaller than decoder sent in SETTINGS).

Best regards,
Tatsuhiro Tsujikawa
=


=C2=A0
Best regards,

Bence B=C3=A9ky


On Fri, Oct 23, 2015 at 12:15 PM, Ilar= i Liusvaara <ilariliusvaara@welho.com> wrote:
On Sat, Oct 24, 2015 at 12:45:49AM +0900, = Tatsuhiro Tsujikawa wrote:
> Hi,
>
> On Fri, Oct 23, 2015 at 1:18 AM, Herv=C3=A9 Ruellan <herve.ruellan@crf.canon.f= r>
> wrote:
>
> > I agree that the wording is ambiguous here.
> >
> > However, my reading is the same a Cory's: you don't have = to send a dynamic
> > table update if the *actual* value is not changed.
> >
> >
> =E2=80=8BI also found the discussion in this ML indicating you are rig= ht.=C2=A0 Thank
> you for clarification.
> I have to ask one more question: what is *actual* value? Is it the tab= le
> size both peer agreed before reading SETTINGS, or the value in
> SETTINGS_HEADER_TABLE_SIZE decoder sent?
>
> I think this is a good item to add in FAQ section..

The way negotiation works:
- Decoder side sets the upper bound via SETTINGS_HEADER_TABLE_SIZE.
- Encoder side sets the actual size via dynamic table updates (inside
=C2=A0 HPACK bitstream) within limits set by decoder.
- If between headers decoder reduces the limit below size signaled by
=C2=A0 encoder, the encoder must first reduce the table size to the minimum=
=C2=A0 it was between the frames or less (it can then increase it up to
=C2=A0 current limit).

As example of the last point:
[4k dynamic table size in use]
--> HEADERS
<-- SETTINGS(SETTINGS_HEADER_TABLE_SIZE=3D4k)
<-- SETTINGS(SETTINGS_HEADER_TABLE_SIZE=3D2k)
<-- SETTINGS(SETTINGS_HEADER_TABLE_SIZE=3D4k)
<-- SETTINGS(SETTINGS_HEADER_TABLE_SIZE=3D8k)
<-- SETTINGS(SETTINGS_HEADER_TABLE_SIZE=3D6k)
--> HEADERS

The second HEADERS must first reduce the dynamic table to at most
2k. It can then increase dynamic table size to up to 6k.


-Ilari




--94eb2c03ed54b052b1053a5916dd--