From mboxrd@z Thu Jan 1 00:00:00 1970 From: Rick Jones Subject: Re: ANNOUNCE: debloat-testing kernel git tree Date: Thu, 03 Mar 2011 10:16:08 -0800 Message-ID: <1299176168.2157.14.camel@tardy> References: <20110225222210.GA3618@tuxdriver.com> <87fwr9jxya.fsf@cruithne.co.teklibre.org> <8739n9ii7z.fsf@cruithne.co.teklibre.org> <87tyfph2az.fsf@cruithne.co.teklibre.org> Reply-To: rick.jones2@hp.com Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: sedat.dilek@gmail.com, linux-wireless@vger.kernel.org, netdev@vger.kernel.org, bloat-devel@lists.bufferbloat.net, linux-kernel@vger.kernel.org To: Dave =?ISO-8859-1?Q?T=E4ht?= Return-path: Received: from g5t0008.atlanta.hp.com ([15.192.0.45]:45325 "EHLO g5t0008.atlanta.hp.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1758616Ab1CCSQQ (ORCPT ); Thu, 3 Mar 2011 13:16:16 -0500 In-Reply-To: <87tyfph2az.fsf@cruithne.co.teklibre.org> Sender: netdev-owner@vger.kernel.org List-ID: > For wireless routers and cable home gateways especially, this research > shows that the total un-managed buffers in your system should be less > than 32. Would it be a good thing to start describing these queues not so much in terms of packets but in terms of delay (or bandwidth X delay)? Constants tend to live longer than they should, no matter how noble in birth. rick jones