W3C home > Mailing lists > Public > ietf-http-wg@w3.org > July to September 2009

Re: p1-message-07 S 7.1.4

From: Henrik Nordstrom <henrik@henriknordstrom.net>
Date: Mon, 20 Jul 2009 10:10:31 +0200
To: Adrien de Croy <adrien@qbik.com>
Cc: Mark Nottingham <mnot@mnot.net>, HTTP Working Group <ietf-http-wg@w3.org>
Message-Id: <1248077431.15530.21.camel@localhost.localdomain>
mån 2009-07-20 klockan 18:08 +1200 skrev Adrien de Croy:

> * a busy - try again later status (like SMTP 421) (actually this would 
> be exceedingly useful for rate limiting connections - a try again with a 
> retry-after header - maybe new status 309?).

503 + Retry-After exists already for this purpose and closely related

3xx is inappropriate as it's not a redirect to a new resource.

4xx is inappropriate as the request as such (when isolated) is fine.

2xx is inappropriate as it's not a successful response.

> * an advertisement for the number of connections a server will accept 
> from a client


> * something else

Not sure anything else is needed.

> However I think these could risk over-complicating the issue for dubious 
> benefit.  Are we even clear on what it is desired to limit?

For normal pageloads with pipelining enabled 2 is a very reasonable

Without pipelining at least 4, maybe more.

For other applications other requrements may exists.

> I've seen this many many times.  Even accessing an idle test server 
> (over GB LAN) via a proxy was hugely slowed down loading many images due 
> to the 2 connection limit in IE7.  It makes sites slow, whether the 
> sites are overloaded or idle. 

Pipelining disabled?

Received on Monday, 20 July 2009 08:11:14 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 11:10:50 UTC