W3C home > Mailing lists > Public > ietf-http-wg@w3.org > July to September 2014

Re: h2 proxy and connection flow control

From: Tatsuhiro Tsujikawa <tatsuhiro.t@gmail.com>
Date: Sat, 23 Aug 2014 00:35:58 +0900
Message-ID: <CAPyZ6=+mft5ACmbevMPb8jgUb0-Maut_bJ6Zww2SzbXLLr5SnA@mail.gmail.com>
To: "ietf-http-wg@w3.org" <ietf-http-wg@w3.org>
On Fri, Aug 22, 2014 at 1:42 PM, Tatsuhiro Tsujikawa <tatsuhiro.t@gmail.com>

> 2014/08/22 0:53 "Amos Jeffries" <squid3@treenet.co.nz>:
> ​​
> >
> > Proxies have to operate under the assumption that a number of other
> > clients and/or streams will be arriving very soon and need to use
> > resources. So no resource can be allocated beyond what can be passed on
> > quickly.
> This is great hint.  The failure I made is always update stream and
> connection window at the same time. It should be updated independently. So
> when Frontend Proxy sends data to Browser, it only sends connection level
> WINDOW_UPDATE to Backend.  Stream level WINDOW_UPDATE is deferred until
> Browser sends strean level WINDOW_UPDATE for data which is not acked.
> Obviously this requires that Frontend Proxy's stream window is not larger
> than Browser's.
Replying to myself, this does not work, because Browser may not send
WINDOW_UPDATE if its stream window is very large (e.g. 10M) compared to
Proxies (e.g., 64K).  Most of the http2 implementation seems to send
WINDOW_UPDATE when window is reduced its half, so sending 64K data does not
make Browser send WINDOW_UPDATE...
​Best regards,
Tatsuhiro Tsujikawa
Received on Friday, 22 August 2014 15:36:46 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 30 March 2016 09:57:10 UTC