From owner-freebsd-current Sat Feb 22 19:50:28 2003 Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 8FB6937B401 for ; Sat, 22 Feb 2003 19:50:26 -0800 (PST) Received: from alpha.siliconlandmark.com (alpha.siliconlandmark.com [209.69.98.4]) by mx1.FreeBSD.org (Postfix) with ESMTP id A612843F93 for ; Sat, 22 Feb 2003 19:50:25 -0800 (PST) (envelope-from andy@siliconlandmark.com) Received: from alpha.siliconlandmark.com (localhost [127.0.0.1]) by alpha.siliconlandmark.com (8.12.6/8.12.6) with ESMTP id h1N3nef0049385; Sat, 22 Feb 2003 22:49:40 -0500 (EST) (envelope-from andy@siliconlandmark.com) Received: from localhost (andy@localhost) by alpha.siliconlandmark.com (8.12.6/8.12.6/Submit) with ESMTP id h1N3neUE049382; Sat, 22 Feb 2003 22:49:40 -0500 (EST) (envelope-from andy@siliconlandmark.com) X-Authentication-Warning: alpha.siliconlandmark.com: andy owned process doing -bs Date: Sat, 22 Feb 2003 22:49:40 -0500 (EST) From: Andre Guibert de Bruet To: Giorgos Keramidas Cc: Erik Greenwald , current@FreeBSD.ORG Subject: Re: SSH (TCP?) lag In-Reply-To: <20030223025707.GE88377@gothmog.gr> Message-ID: <20030222220054.C34711@alpha.siliconlandmark.com> References: <20030222100441.Y34711@alpha.siliconlandmark.com> <20030222125817.A15590@xarx.midsouth.rr.com> <20030222200117.E34711@alpha.siliconlandmark.com> <20030223025707.GE88377@gothmog.gr> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG On Sun, 23 Feb 2003, Giorgos Keramidas wrote: > On 2003-02-22 20:05, Andre Guibert de Bruet wrote: > > > I noticed the same thing... then > > > > > > try sysctl net.inet.tcp.delayed_ack=0 > > > > > > fixed the issue > > > > That worked. Shouldn't this sysctl be turned off by default? > > Nah. Not really. Delaying acks can save quite a lot of of bandwidth > for bulk data transfers. Having read up on the issue, I can understand the reasoning for wanting delayed_ack on by default. From tuning(7): With delayed acks turned off, the acknowledgement may be sent in its own packet, before the remote service has a chance to echo the data it just received. This same concept also applies to any interactive protocol (e.g. SMTP, WWW, POP3), and can cut the number of tiny packets flowing across the network in half. The FreeBSD delayed ACK implementation also follows the TCP protocol rule that at least every other packet be acknowledged even if the standard 100ms timeout has not yet passed. Normally the worst a delayed ACK can do is slightly delay the teardown of a connection, or slightly delay the ramp-up of a slow- start TCP connection. I find myself waiting up to two seconds for data to flush to the terminal on a 28 line 'ls -l'. net.inet.tcp.delayed_ack doesn't appear to cause this behavior on 4.7-stable. Did we inadvertently break the 100ms clause with the latest TCP patches? Regards, > Andre Guibert de Bruet | Enterprise Software Consultant > > Silicon Landmark, LLC. | http://siliconlandmark.com/ > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message