From owner-freebsd-net Sat Apr 6 6:57:43 2002 Delivered-To: freebsd-net@freebsd.org Received: from iguana.icir.org (iguana.icir.org [192.150.187.36]) by hub.freebsd.org (Postfix) with ESMTP id 36C6637B404 for ; Sat, 6 Apr 2002 06:57:40 -0800 (PST) Received: (from rizzo@localhost) by iguana.icir.org (8.11.6/8.11.3) id g36EvaR29466; Sat, 6 Apr 2002 06:57:36 -0800 (PST) (envelope-from rizzo) Date: Sat, 6 Apr 2002 06:57:36 -0800 From: Luigi Rizzo To: Andreas Klemm Cc: freebsd-net@freebsd.org Subject: Re: better DSL bandwidth usage by priorizing ACKs in outgoing packets over others Message-ID: <20020406065735.A29144@iguana.icir.org> References: <20020402163912.GD4307@titan.klemm.gtn.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20020402163912.GD4307@titan.klemm.gtn.com> User-Agent: Mutt/1.3.23i Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Tue, Apr 02, 2002 at 06:39:12PM +0200, Andreas Klemm wrote: we would need a minor tweak to the ipfw code so that it can match packets whose size is less than X bytes (so the mechanism is general enough to be used for other things). This could be done in a matter of 1hour or less, most of the time would be wasted in figuring out a way to implement it that does not break binary compatibility. Once we have done this, we can define a dummynet pipe with a bandwidth <= the bottleneck (128kbit/s), and use"queue" rules to privilege the acks wrt other traffic. Something like ipfw pipe 10 config bw 100kbit/s ipfw queue 1 config weight 1 pipe 10 ipfw queue 100 config weight 100 pipe 10 ipfw add queue 100 tcp from any to ${outside} shorter-than 80 ipfw add queue 1 ip from any to ${outside} This said, i have never seen terribly bad effects when cvsupping and doing other things. If there is something which goes to its knees, this is the disk. Oh, and "usually" in italy, the cheap DSL contracts give you 256kbit/s in! cheers luigi > Hi ! > > A collegue of mine has an Apple (Mac OS X) and told me about > a cool software, that priorizes outgoing ACKs over other traffic. > > On Tue, Apr 02, 2002 at 04:53:08PM +0200, andreas.klemm.ak@bayer-ag.de wrote: > > http://www.intrarts.com/quest/throttle.html > > Using DSL you have usually 768K in and 128K out. > Figure a szenario, where you ftp like hell from a ftp > server. You get very good throughput, since the outgoing > ACKs of the incoming ftp data stream are not throttled by > 128K outgoing bandwidth. > > But if you start another application like cvsup at the > same time you'll notice an immediate throttle of incoming > packets, because cvsup monopolizes the outgoing 128K bandwidth. > And if the ACKs don't get out in a timely manner, you can't get > that much incoming FTP traffic. > > The above mentioned software manages exactly this by using > a daemon program. With divert sockets and ipfw you send outgoing > traffic to this daemon which gives outgoing ACKs over the > 128K higher preference and buffers the rest of the traffic > (I think). > > Would something like this be possible anyhow with our current > Firewall implementation, or would somebody have time and fun > to implement this ?? > > Andreas /// > > -- > Andreas Klemm /\/\/\/\/\/\/\/\/\/\/\ > http://www.64bits.de < Powered by FreeBSD > > http://www.apsfilter.org/ \ www.FreeBSD.org / > http://people.FreeBSD.ORG/~andreas \/\/\/\/\/\/\/\/\/\/ > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message