From owner-freebsd-fs@FreeBSD.ORG  Mon Jul 24 15:28:24 2006
Return-Path: <owner-freebsd-fs@FreeBSD.ORG>
X-Original-To: fs@freebsd.org
Delivered-To: freebsd-fs@FreeBSD.ORG
Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125])
	by hub.freebsd.org (Postfix) with ESMTP id 8738E16A4DA
	for <fs@freebsd.org>; Mon, 24 Jul 2006 15:28:24 +0000 (UTC)
	(envelope-from rees@citi.umich.edu)
Received: from citi.umich.edu (citi.umich.edu [141.211.133.111])
	by mx1.FreeBSD.org (Postfix) with ESMTP id 5443843D45
	for <fs@freebsd.org>; Mon, 24 Jul 2006 15:28:24 +0000 (GMT)
	(envelope-from rees@citi.umich.edu)
Received: from citi.umich.edu (dumaguete.citi.umich.edu [141.211.133.51])
	by citi.umich.edu (Postfix) with ESMTP id C32D81BB93;
	Mon, 24 Jul 2006 11:28:23 -0400 (EDT)
From: Jim Rees <rees@umich.edu>
To: rick@snowhite.cis.uoguelph.ca
In-Reply-To: <200607241456.KAA63411@snowhite.cis.uoguelph.ca>
	(rick@snowhite.cis.uoguelph.ca, Mon, 24 Jul 2006 10:56:50 EDT)
Date: Mon, 24 Jul 2006 11:28:23 -0400
Sender: rees@citi.umich.edu
Message-Id: <20060724152823.C32D81BB93@citi.umich.edu>
Cc: fs@freebsd.org
Subject: Re: freebsd4.11 patch for nfs over tcp 
X-BeenThere: freebsd-fs@freebsd.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Filesystems <freebsd-fs.freebsd.org>
List-Unsubscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-fs>,
	<mailto:freebsd-fs-request@freebsd.org?subject=unsubscribe>
List-Archive: <http://lists.freebsd.org/pipermail/freebsd-fs>
List-Post: <mailto:freebsd-fs@freebsd.org>
List-Help: <mailto:freebsd-fs-request@freebsd.org?subject=help>
List-Subscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-fs>,
	<mailto:freebsd-fs-request@freebsd.org?subject=subscribe>
X-List-Received-Date: Mon, 24 Jul 2006 15:28:24 -0000

Delayed acks when properly implemented should not cause a performance hit.
The problem is that there are many buggy implementations.  If you have to
turn off delayed acks to get good nfs performance, then there is a bug in
the tcp stack.  See, for example, rfc 2525 page 40, "Stretch ACK violation."
MacOS has this same problem.

Having said that, the code is obviously wrong.  It's trying to set NODELAY
and failing.  This has apparently been fixed, because 6.1 does not have this
bug.  I'm pretty sure 4.x is out of maintenance.