From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail.mediaspot.net (mail.mediaspot.net [109.73.92.181]) by huchra.bufferbloat.net (Postfix) with ESMTP id 1BF7721F15E for ; Thu, 24 Jan 2013 09:50:01 -0800 (PST) X-Spam-Status: No, hits=4.1 required=6.5 tests=AWL: 0.125,BAYES_99: 4.07,HTML_MESSAGE: 0.001, TOTAL_SCORE: 4.196,autolearn=no X-Spam-Level: **** X-Footer: bWVkaWFzcG90Lm5ldA== Received: from exchange.mediaspot.net ([192.168.0.22]) by mail.mediaspot.net (using TLSv1/SSLv3 with cipher AES128-SHA (128 bits)) for Codel@lists.bufferbloat.net; Thu, 24 Jan 2013 18:49:57 +0100 Received: from EXCHANGE.mediaspot.local ([fe80::59f7:a31d:de3f:869b]) by EXCHANGE.mediaspot.local ([fe80::59f7:a31d:de3f:869b%26]) with mapi id 14.02.0328.009; Thu, 24 Jan 2013 18:49:57 +0100 From: Alessandro Bolletta To: "Codel@lists.bufferbloat.net" Thread-Topic: In which terms ath9k queuing limits fq_codel? Thread-Index: AQHN+lrd/+SjS3FXtkihhRYhuBMUCQ== Date: Thu, 24 Jan 2013 17:49:56 +0000 Message-ID: Accept-Language: it-IT, en-US Content-Language: it-IT X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [192.168.0.125] Content-Type: multipart/alternative; boundary="_000_F52F175DFC537F48A43937B50B44BCA276D305A6EXCHANGEmediasp_" MIME-Version: 1.0 Subject: [Codel] In which terms ath9k queuing limits fq_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: Thu, 24 Jan 2013 17:50:02 -0000 --_000_F52F175DFC537F48A43937B50B44BCA276D305A6EXCHANGEmediasp_ Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi, I have a doubt about the problem that fq_codel has with 802.11n packet aggr= egation, and your help to understand the problem would be well accepted. Could you explain me how and in which real terms (for example, through an e= xample) ath9k underlaying queue will limit fq_codel effect? I imagine that ath9k will only add some delay that can't be controlled by f= q_codel, but I can't imagine how delay could grow in case of fillfullment o= f ath9k's queue... Also, I read that fq_codel drops packets at head of queue. So, if fq_codel = can't manage and act over ath9k queue, will fq_codel's effect change and no= t to work as expected? I'm thinking, for example, at the most common TCP co= ngestion control techniques: will TCP be able to correctly understand what = fq_codel wanted to announce, if it isn't droping packets at the very first = level of queuing? Thanks Alessandro Bolletta --_000_F52F175DFC537F48A43937B50B44BCA276D305A6EXCHANGEmediasp_ Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

Hi,
I have a doubt about the problem that fq_codel has with 802.11n packet aggr= egation, and your help to understand the problem would be well accepted.
Could you explain me how and in which real terms (for example, through an e= xample) ath9k underlaying queue will limit fq_codel effect?
I imagine that ath9k will only add some delay that can't be controlled by f= q_codel, but I can't imagine how delay could grow in case of fillfullment o= f ath9k's queue...

Also, I read that fq_codel drops packets at head of queue. So, if fq_codel = can't manage and act over ath9k queue, will fq_codel's effect change and no= t to work as expected? I'm thinking, for example, at the most common TCP co= ngestion control techniques: will TCP be able to correctly understand what fq_codel wanted to announce, if i= t isn't droping packets at the very first level of queuing?

Thanks

Alessandro Bolletta

--_000_F52F175DFC537F48A43937B50B44BCA276D305A6EXCHANGEmediasp_--