From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-wg0-f47.google.com (mail-wg0-f47.google.com [74.125.82.47]) (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 F01BE208AB0 for ; Sun, 20 May 2012 20:48:37 -0700 (PDT) Received: by wgbfa7 with SMTP id fa7so3718588wgb.28 for ; Sun, 20 May 2012 20:48:34 -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=CVqotSe+64NK9rM48ln9iSHKolLqrdPoPtXNYIQDi9Y=; b=rtgFE8NVCrC1+mYGaKyyiyOubeAnJBnpYkeDGzLYSeiNsHp4WBxF2QFt2A66e4Q+g3 WIQZzGtVRtjffhIZ8SYHbMhjJei3Om7YEm27BvIRibyIjgLrysFjIp0VHlhyxe82WDMY qN1xJZDppWBKAcvbdnzX2MyBTQYK/z8RRoqmCbApL5iohpoVuOTB1DFgO7lnhjjPllxj XDdCB1lBivDIgWu8zLv3eLof8AnMprixMonn9s4WX5qZCIorQffyMmenBokR68fws7Oy 21sdMQqCbsmumXXLBE6ZWJ6e5VCkEQOAJu/CqLrp3uLGqncKYBzRb/yGu6Lnw0AAh8DL FW4A== MIME-Version: 1.0 Received: by 10.180.88.194 with SMTP id bi2mr21095678wib.20.1337572114707; Sun, 20 May 2012 20:48:34 -0700 (PDT) Received: by 10.223.115.9 with HTTP; Sun, 20 May 2012 20:48:34 -0700 (PDT) In-Reply-To: <20120521003115.GO22418@yumi.tdiedrich.de> References: <20120520212944.GK22418@yumi.tdiedrich.de> <20120521003115.GO22418@yumi.tdiedrich.de> Date: Mon, 21 May 2012 04:48:34 +0100 Message-ID: From: Dave Taht To: Tobias Diedrich Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Cc: OpenWrt Development List , codel@lists.bufferbloat.net Subject: Re: [Codel] BQL support in Ethernet drivers (and Kathie Nichols and Van Jacobson's new AQM, codel) 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, 21 May 2012 03:48:38 -0000 Thx for the numbers! Could you do a TCP_RR while under load from UDP_STREAM? On Mon, May 21, 2012 at 1:31 AM, Tobias Diedrich wrote: > Tobias Diedrich wrote: >> Dave Taht wrote: >> > In looking over the enormous stack of boards and drivers that openwrt >> > supports, I see that many of the ethernet drivers don't yet support >> > Linux 3.3's "Byte Queue Limits", which are discussed here: >> > >> > http://lwn.net/Articles/454390/ >> > >> > It would be good if more did. They improve network performance in the >> > general case enormously, particularly when a link is not connected at >> > it's peak wire speed. >> > >> > *Adding* support for BQL to an ethernet driver is trivial, here's an >> > example of how. >> >> I tried adding BQL to the ramips ethernet driver, however I found >> some interesting behaviour while doing >> "root@OpenWrt:~# netperf -l 120 -t UDP_STREAM -H myserver" >> >> It looks like the briding code still needs to implement this as well? >> >> netperf UDP_STREAM: >> iface =A0limit_min =A0 inflight =A0tx mbps =A0remote mbps =A0ping ms >> eth0 =A0 0 =A0 =A0 =A0 =A0 =A0 ~15000 =A0 =A095.71 =A0 =A095.71 =A0 =A0 = =A0 =A0~10ms >> eth0 =A0 1000000 =A0 =A0 ~300000 =A0 177.98 =A0 23.28(*) =A0 =A0 ~30ms >> br0 =A0 =A00 =A0 =A0 =A0 =A0 =A0 ~15000 =A0 =A0154.88 =A0 33.94(*) =A0 = =A0 ~120ms >> br0 =A0 =A01000000 =A0 =A0 ~300000 =A0 170.92 =A0 25.57(*) =A0 =A0 ~30ms >> >> (*) bwm-ng on the server showed ~100mbps incoming... > [...] >> Haven't tried codel yet... > > Turns out, it works nicely with codel, even with the bridge: > > netperf: =A0netperf -l 120 -t UDP_STREAM -H myserver > fq_codel: tc qdisc add dev eth0 handle 1: root fq_codel target 5ms > > iface =A0eth0 qdisc =A0bql =A0inflight =A0tx mbps =A0 =A0sys time =A0ping= ms > eth0 =A0 pfifo_fast =A0no =A0 n/a =A0 =A0 =A0 182.98(*) =A096.43s =A0 =A0= ~30ms > eth0 =A0 fq_codel =A0 =A0no =A0 n/a =A0 =A0 =A0 177.98(*) =A096.09s =A0 = =A0~30ms > eth0 =A0 pfifo_fast =A0yes =A0~15000 =A0 =A095.71 =A0 =A0 =A042.73s =A0 = =A0~10ms > eth0 =A0 fq_codel =A0 =A0yes =A0~15000 =A0 =A095.19 =A0 =A0 =A051.52s =A0= =A0~4ms > br0 =A0 =A0pfifo_fast =A0yes =A0~15000 =A0 =A0155.19(*) =A094.24s =A0 =A0= ~120ms > br0 =A0 =A0fq_codel =A0 =A0yes =A0~15000 =A0 =A090.92 =A0 =A0 =A065.52s = =A0 =A0~4ms > > (*) 100mbit link after the switch, ifconfig eth0 shows no drops, > =A0 =A0so I'm assuming they are getting dropped by the switch. > > -- > Tobias =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 = =A0 =A0 =A0 =A0 =A0PGP: http://8ef7ddba.uguu.de --=20 Dave T=E4ht SKYPE: davetaht US Tel: 1-239-829-5608 http://www.bufferbloat.net