From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mout.gmx.net (mout.gmx.net [212.227.15.19]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client CN "mout.gmx.net", Issuer "TeleSec ServerPass DE-1" (verified OK)) by huchra.bufferbloat.net (Postfix) with ESMTPS id 14EB521F454 for ; Wed, 3 Jun 2015 09:01:24 -0700 (PDT) Received: from [10.185.27.36] ([46.189.28.54]) by mail.gmx.com (mrgmx002) with ESMTPSA (Nemesis) id 0LmKOI-1ZYxtF1oqg-00Zsvr; Wed, 03 Jun 2015 18:01:20 +0200 User-Agent: K-9 Mail for Android In-Reply-To: References: MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----D5O0ZGB1243UPV8STGNGFV8UF7LWO7" Content-Transfer-Encoding: 7bit From: Sebastian Moeller Date: Wed, 03 Jun 2015 17:58:03 +0200 To: Aaron Wood , cerowrt-devel Message-ID: X-Provags-ID: V03:K0:KCjA+5zzwhdDRX6HhEOrQxDnbqbk+Eq0HgPFGlJudjQdRSZ4Kug gwzZugYejRNzBw3VKoYIzSvOU53u+4deJD55LfVNh4oNrNmWg1z/eTvN6HWIglDIYG8CpoJ 6nA07aRNrEUq/aGLFZhT9b5fHC2c9GaRVyAitwVBeq4vRfvWyD0Np3LOsK9JSROdQo4eHnS m/wZFkOfYWwuwN/ahLFFw== X-UI-Out-Filterresults: notjunk:1;V01:K0:7THWi3zlBXM=:8JIvWnT45wjVIV4CrVVidn lZYKs4k7aXxK/+EztZ2LkUjVkkLuWWFlAoaU2JN0HzR4E2P8T+BiGX0sU1+YUgsg2ZbAm+bSj Hpw+yytnv/xe6n9eaqrFtb9aFcT8HIk5MD0jq6oEMQskSA5W4vK7J31IVLdcXJ2AoXZ34qK4y FvcduY+dKw8VmLAbzdCKWOfQ7rEyUhJtGfSo9dScoC6dlZOPwPVVLkg6iz8IucvirC5JV/UAS eRos4jC0iDVsDySdfukXBjnv/0rjIX6edgbijRIix6s62fwlpayCVv/chVynbQJCCYYRD2vL9 nFaC9NapzKM7vKflcnem/gQ7N7+8eW6+zvLV+Gb6kh5ibsa3M8qqMjKnCTvyY4UawKO5odmzq CD+L2FBsWJUGVmD3RACUiuXo2x5gbgIqZfJt2eXBtkfcHkEmDjrmo4C3wWXcZQ8bDCbD4x7K+ UhH/E8Ufj3QyI4VmPnCznR7wgS34hzo2GbJYa+Q3gQfhROlUQdO7JSqzw6wghk30j9oVTU51h 5kkYX6UdTS7S5trRH5lcToZOOuojMUORR/5FeKgMjain3p3xDD6NDx0ZFs4i4D+SocHrPGYgq p8zn3PFMybBgMMojKEcsZZ6ozWoenLbUyE/rq9iggX+G3h7N8yChvAm3JAOQNOUyYrptR0ndk ctllMx3tcKbQB5iYEyzLG/GEr3TiGGL/SAaPy0IC65SfeX9jmV5mSj/JUQ4Spe5hOgBdtmS1t r4eLxz9kb+eEdluI Subject: Re: [Cerowrt-devel] ingress rate limiting falling short X-BeenThere: cerowrt-devel@lists.bufferbloat.net X-Mailman-Version: 2.1.13 Precedence: list List-Id: Development issues regarding the cerowrt test router project List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 03 Jun 2015 16:01:53 -0000 ------D5O0ZGB1243UPV8STGNGFV8UF7LWO7 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Hi Aaron, about the 5% loss with the wndr, please remember that the shaper works typ= ically on raw Ethernet rates, while flent reports TCP good put I believe=2E= So roughly 2 to 6 percent difference can be explained with a combination o= f the following overheads: PTM/ATM, ethernet, VLAN(s), PPPoE, IPv4/IPv6, TC= P, potential TCP options like timestamps=2E=2E=2E Now flent might report ac= tual tcp-rates and I am out to lunch, but I have a hard time reconciling h= ow flent/netperf, can actually learn about all the additional overhead=2E= =2E=2E As far as I can tell all netperf sees is the payload it sends and th= e payload it receives=2E=2E=2E Tl;dr: Overhead unknown to flent effortlessly explains why the TCP good pu= t as reported by flent falls short of the shaped rate as that rate contains= all overheads=2E Best Regards Sebastian On June 3, 2015 7:45:47 AM GMT+02:00, Aaron Wood wro= te: >I wrote this up on my blog, where I can intersperse text and graphs a >bit >better: > >http://burntchrome=2Eblogspot=2Ecom/2015/06/htb-rate-limiting-not-quite-l= ining-up=2Ehtml > >Basically, I ran a series of tcp_download tests, using increasing >ingress >rates with sqm_scripts, and then used flent's box-plots to put the >results >into a combined image for comparing=2E > >On the 3800, it never meets the rate, but it's only off by maybe 5%=2E=20 >But >on my new WRT1900AC, it's wildly off, even over the same performance >range >(I tested it from 80-220Mbps rates in 20Mbps jumps, and saw from >40-150Mbps=2E > >I have no idea where to start looking for the cause=2E But for now, I'm >just >setting my ingress rate MUCH higher than I should, because it's working >out >to the right value as a result=2E > >-Aaron > > >------------------------------------------------------------------------ > >_______________________________________________ >Cerowrt-devel mailing list >Cerowrt-devel@lists=2Ebufferbloat=2Enet >https://lists=2Ebufferbloat=2Enet/listinfo/cerowrt-devel --=20 Sent from my Android device with K-9 Mail=2E Please excuse my brevity=2E ------D5O0ZGB1243UPV8STGNGFV8UF7LWO7 Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: quoted-printable Hi Aaron,

about the 5% loss with the wndr, please remember that the shaper works typ= ically on raw Ethernet rates, while flent reports TCP good put I believe=2E= So roughly 2 to 6 percent difference can be explained with a combination o= f the following overheads: PTM/ATM, ethernet, VLAN(s), PPPoE, IPv4/IPv6, TC= P, potential TCP options like timestamps=2E=2E=2E Now flent might report ac= tual tcp-rates and I am out to lunch, but I have a hard time reconciling h= ow flent/netperf, can actually learn about all the additional overhead=2E= =2E=2E As far as I can tell all netperf sees is the payload it sends and th= e payload it receives=2E=2E=2E
Tl;dr: Overhead unknown to flent effortlessly explains why the TCP good pu= t as reported by flent falls short of the shaped rate as that rate contains= all overheads=2E


Best Regards
Sebastian

On June 3, 2015 7:45:4= 7 AM GMT+02:00, Aaron Wood <woody77@gmail=2Ecom> wrote:
I wrote this up on my blog, where I can intersperse = text and graphs a bit better:

http://burntchrome=2Eblogspot=2Ecom/2015/06/htb-rate-limiting-not-quite-= lining-up=2Ehtml

Basically, I ran a series o= f tcp_download tests, using increasing ingress rates with sqm_scripts, and = then used flent's box-plots to put the results into a combined image fo= r comparing=2E

On the 3800, it never meets the r= ate, but it's only off by maybe 5%=2E=C2=A0 But on my new WRT1900AC, it= 's wildly off, even over the same performance range (I tested it from 8= 0-220Mbps rates in 20Mbps jumps, and saw from 40-150Mbps=2E

I have no idea where to start looking for the cause=2E=C2=A0 Bu= t for now, I'm just setting my ingress rate MUCH higher than I should, = because it's working out to the right value as a result=2E
-Aaron


Cerowrt-devel mailing list<=
br />Cerowrt-devel@lists=2Ebufferbloat=2Enet
https://lists=2Ebufferbloat=2E=
net/listinfo/cerowrt-devel

--
Sent from my Android device with K-9 Mail=2E Please excuse my brevity=2E ------D5O0ZGB1243UPV8STGNGFV8UF7LWO7--