From owner-freebsd-net  Mon Oct  2  7:33:17 2000
Delivered-To: freebsd-net@freebsd.org
Received: from ptavv.es.net (ptavv.es.net [198.128.4.29])
	by hub.freebsd.org (Postfix) with ESMTP id DB20D37B502
	for <net@FreeBSD.ORG>; Mon,  2 Oct 2000 07:33:14 -0700 (PDT)
Received: from ptavv.es.net (localhost [127.0.0.1])
	by ptavv.es.net (8.10.1/8.10.1) with ESMTP id e92EX6r15583;
	Mon, 2 Oct 2000 07:33:06 -0700 (PDT)
Message-Id: <200010021433.e92EX6r15583@ptavv.es.net>
To: Jonathan Lemon <jlemon@flugsvamp.com>
Cc: julian@elischer.org, net@FreeBSD.ORG
Subject: Re: SACK in FreeBSD TCP. 
In-reply-to: Your message of "Mon, 02 Oct 2000 08:58:37 CDT."
             <200010021358.e92DwbA14157@prism.flugsvamp.com> 
Date: Mon, 02 Oct 2000 07:33:06 -0700
From: "Kevin Oberman" <oberman@es.net>
Sender: owner-freebsd-net@FreeBSD.ORG
Precedence: bulk
X-Loop: FreeBSD.org

I can't comment on what the symptoms are in Oz, but testing on very
large streams done last year by ESnet, Lawrence Berkeley National
Laboratory, and Argonne National Laboratory made it painfully obvious
that trying to run a single TCP stream of over 200 Mbps between
Chicago and California was impossible without SACK. It does require
many other things, like very large windows and fast recovery, but
without SACK it's hopeless.

In our testing we used Solaris and Tru64 UNIX, but FreeBSD was not an
option because of the lack of SACK. As a strong proponent of FreeBSD
and the FreeBSD network stack in particular, this was a bit painful to
me.

R. Kevin Oberman, Network Engineer
Energy Sciences Network (ESnet)
Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab)
E-mail: oberman@es.net			Phone: +1 510 486-8634


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-net" in the body of the message