W3C home > Mailing lists > Public > ietf-http-wg@w3.org > October to December 2002

Re: Connection:Keep-Alive and Proxies

From: Jeffrey Mogul <Jeff.Mogul@hp.com>
Date: Wed, 27 Nov 2002 11:20:16 -0800
Message-Id: <200211271920.gARJKGgD028492@wera.hpl.hp.com>
To: Alex Rousskov <rousskov@measurement-factory.com>
cc: Diwakar Shetty <Diwakar.Shetty@oracle.com>, ietf-http-wg@w3.org

Alex wrote:

   I think the book is misquoting the RFC.

Actually, I think Diwakar Shetty's original message misquoted the book.
(Alex guess this right.)  He wrote:

    Following is a para which i read in one book.

    An HTTP/1.0 client could send a "Keep-Alive" header to a HTTP/1.1
    proxy that did not understand "Connection" but might mistakenly
    forward it. If the downstream connection also maintained a
    "Keep-Alive" connection, the proxy in the middle would never
    receive the closing of the response

He later identified this as from page 289 in Krishnamurthy & Rexford.
The actual paragraph is much longer (the quote above is not the
whole paragraph!) and what the paragraph in the book actually says is:

	[2 sentences I'm not quoting]
	However, interaction between the Connection header and
	Keep-Alive header could result in a hung connection.  This
	occured because an HTTP/1.0 client could send a "Keep-Alive"
	header to a proxy that did not understand "Connection" but
	might mistakenly forward it.  If the downstream connection also
	maintained a "Keep-Alive" connection, the proxy in the middle
	would never receive the closing of the response.  To avoid such
	problems, HTTP/1.1 proxies are not permitted to establish a
	persistent connection with HTTP/1.0 clients.

So the discussion on this mailing list has been misguided because
the book never mentioned "a HTTP/1.1 proxy that did not understand
'Connection'".  

I'm sure the Krishnamurthy & Rexford book does have bugs, and
I suspect this paragraph could have been clearer if the phrase
"to a proxy" had been "to an HTTP/1.0 proxy".  But from now on,
let's insist on accurate quotes before discussing whether some
publication got the story right.

-Jeff
Received on Wednesday, 27 November 2002 14:20:22 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 06:49:21 GMT