From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-pz0-f43.google.com (mail-pz0-f43.google.com [209.85.210.43]) (using TLSv1 with cipher RC4-MD5 (128/128 bits)) (Client CN "smtp.gmail.com", Issuer "Google Internet Authority" (verified OK)) by huchra.bufferbloat.net (Postfix) with ESMTPS id 119B321F0E2 for ; Mon, 14 May 2012 11:31:16 -0700 (PDT) Received: by dajz8 with SMTP id z8so10206994daj.16 for ; Mon, 14 May 2012 11:31:15 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type:content-transfer-encoding; bh=5ClUAMcbfU/ttjWCTA40SZ9tSvvlapcexX15ilYC8dk=; b=ujq5ZFjtfDOHwUrTf5qwi5qB2suf6PejDAtjo8+kBRjgJWdgGuErwGrSa5C+JjVR/3 xL0x4NKb1vNUMh6AkQDq/GTXG6baTtkXJG1m8NF0r2KB1xch9O+qLIwICH9J0Ks9t4gF waNXlxUY+qwcmUQoDVmSxFTI6DKZzY6ICClrSURUjhJUP8vJvQ7We+MbnlpylOJy5y8S s/IBcVgL5OgAEJxnosxIrR4u+b3cid+hGZl3CfQ2N/qdB02OxFyxRYJAv7r+fKBvD/qI byXT7KB7+2DTrwTC65ggaxABroP4nEh/EUqoZs4BD5SDYwZOIxlx9nBAJq+/QMc/VbLv r+iA== MIME-Version: 1.0 Received: by 10.50.195.234 with SMTP id ih10mr179445igc.0.1337020275197; Mon, 14 May 2012 11:31:15 -0700 (PDT) Received: by 10.64.94.129 with HTTP; Mon, 14 May 2012 11:31:15 -0700 (PDT) In-Reply-To: References: <1336855256.31653.1329.camel@edumazet-glaptop> <20120512.164513.1156706853054390966.davem@davemloft.net> <1336859324.31653.1385.camel@edumazet-glaptop> <20120512.175217.1632102067268101115.davem@davemloft.net> <1336893803.8512.43.camel@edumazet-glaptop> <4FB09F8D.6070403@gmail.com> <1336976277.8512.46.camel@edumazet-glaptop> <4FB0A73F.9090501@gmail.com> <4FB0AA68.2000800@gmail.com> <1336983815.8512.89.camel@edumazet-glaptop> Date: Mon, 14 May 2012 20:31:15 +0200 Message-ID: From: =?ISO-8859-1?Q?Roger_J=F8rgensen?= To: Dave Taht Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Cc: codel@lists.bufferbloat.net Subject: Re: [Codel] [PATCH net-next] codel: use Newton method instead of sqrt() and divides X-BeenThere: codel@lists.bufferbloat.net X-Mailman-Version: 2.1.13 Precedence: list List-Id: CoDel AQM discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 14 May 2012 18:31:16 -0000 On Mon, May 14, 2012 at 5:05 PM, Dave Taht wrote: > On Mon, May 14, 2012 at 4:34 AM, Roger J=F8rgensen wro= te: >> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 rtt min =A0 =A0avg =A0 = =A0 max =A0 =A0 mdev =A0 =A0bw test showing >> plain =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A01,037 =A0 =A0 =A02,327 =A05,123= =A0 0,558 =A0 =A02,4Gbps >> TBF =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 1,529 =A0 =A0 =A02,87 =A0 8,275 = =A0 0,547 =A0 =A01,5Gbps >> fq_codel +ecn =A0 =A0 0,907 =A0 =A0 1,82 =A0 =A05,214 =A0 0,547 =A0 =A01= ,4Gbps >> fq_codel noec =A0 =A0 0,905 =A0 =A0 1,913 =A0 4,95 =A0 =A00,586 =A0 1,4G= bps > > > I find your results mildly puzzling. certainly your max rtt in the no > ecn case is pleasing, > but your max throughput in all three AQM cases seems to be rather > severely capped. I did try to capp the bandwidth, (that's why the values for fq_codel is as they are), the idea was to limit the bw and see how the pingtime vary with that. Turns out that bw capping give better pingtime than no capping :-) and since I can see how it work I'll try to run it on a linux box acting as gateway for a wireless guestnetwork to see it under real load :-) --=20 Roger Jorgensen=A0 =A0 =A0 =A0 =A0=A0=A0| ROJO9-RIPE rogerj@gmail.com=A0 =A0 =A0 =A0=A0 =A0| - IPv6 is The Key! http://www.jorgensen.no=A0=A0 | roger@jorgensen.no