From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from webmail.newmedia-net.de (smtps.newmedia-net.de [IPv6:2a05:a1c0:0:de::167]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by lists.bufferbloat.net (Postfix) with ESMTPS id 20DE43B2A4 for ; Tue, 20 Aug 2019 14:40:42 -0400 (EDT) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=newmedia-net.de; s=mikd; h=Content-Transfer-Encoding:Content-Type:In-Reply-To:MIME-Version:Date:Message-ID:From:References:Cc:To:Subject; bh=SS7/6pvYw6y6KLkPkzi/bxQOLTatrYuLfCHJ2TtfdoE=; b=A+JmVf1qqT2SmzQHk2K3zOrhX4/dWJE+82eSAKqnWQPEOSKFFV833WS1ZqTN38zNOoSD2+JZdiRbMxtltW866vAYjc1XUXlUowt32bu5nq1D0gpKJCxfD97w1rKkCrGEsHvZVCA9nwESD24KfwnMnI38CeAmWLxIQh0ZFWFz1TQ=; To: =?UTF-8?Q?Toke_H=c3=b8iland-J=c3=b8rgensen?= , Dave Taht Cc: "cake@lists.bufferbloat.net >> Cake List" References: <384866b4-4c91-cf2c-c267-ee4036e5fbf7@newmedia-net.de> <87wof7sriw.fsf@toke.dk> <6782ec15-30eb-63b0-f54f-376c5e6b840b@newmedia-net.de> <87tvabsp99.fsf@toke.dk> <74bccc2b-b805-255f-b6a7-83ade9af6765@newmedia-net.de> <87r25fsn70.fsf@toke.dk> From: Sebastian Gottschall Message-ID: Date: Tue, 20 Aug 2019 20:39:38 +0200 User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:60.0) Gecko/20100101 Thunderbird/60.8.0 MIME-Version: 1.0 In-Reply-To: <87r25fsn70.fsf@toke.dk> Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 8bit X-Received: from [2003:c9:3f25:1300:e947:ab73:1d88:4792] by webmail.newmedia-net.de with esmtpsa (TLSv1:AES128-SHA:128) (Exim 4.72) (envelope-from ) id 1i093n-0003zU-If; Tue, 20 Aug 2019 20:40:47 +0200 Subject: Re: [Cake] cake in dd-wrt X-BeenThere: cake@lists.bufferbloat.net X-Mailman-Version: 2.1.20 Precedence: list List-Id: Cake - FQ_codel the next generation List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 20 Aug 2019 18:40:42 -0000 Am 20.08.2019 um 20:31 schrieb Toke Høiland-Jørgensen: > Sebastian Gottschall writes: > >>>> we are already using filters. yes. its just that cake is acting always >>>> as root and we have different sorts of qos configurations. so you have >>>> wan. but we may have multiple lan interfaces with individual qos >>>> settings. the same for mac / ip based user settings. so in fact we need >>>> to create a individual qdisc for each of these setting types in worst >>>> case, but in that case we cannot take in account the global available >>>> bandwidth anymore. >>> Ah, right, I see. So this is things like users wanting to limit a >>> specific type of traffic to a certain bandwidth? >> basicly yes. there are multiple ways. plain traffic shaping by local >> interface name, by local mac, by local ip/net >> and in addition there is shaping by port based or dpi based packet >> detection and since each instance of cake doesnt know of any other >> use of cake qdiscs its getting complicated. but we just started with >> working on it. i'm sure i find a solution for it > Do let us know if you do :) > > However, I'd also point out that when running CAKE a lot of these kinds > of setups become simply redundant. For home networks most of the setups > I have seen with such rule-based shaping is simply there to paper over > the underlying bufferbloat issue. Once you solve that you don't really > need all the policy-based stuff. its not just about policy to get all managed. but the point is that a heavy bittorrent downloader will still steal the bandwidth of my scp session. so its about control and not just about the flow management is about limiting ports to a specific bandwidth. for instance. i have a concert venue and i limit the backstage network to a certain maximum rate since a need a budged for other networks so i limit the ethernet port of this network on the main router to lets say 10 mbit or something like that priorize torrent and other bad services to bulk. which just works good for internet. so we have enough bandwidth on our other cables for doing 4k streams. dd-wrt is not just used on these plastic routers for home users. this is one option and works great without much qos configuration. you're right. but if its turning more complex and professional its not enough anymore. > Now, there are of course exceptions to this where a strict rule-based > shaping *is* really needed; but HTB already provides this in the kernel, > and we don't want to re-invent that, so I'm not sure we'll ever support > this properly in CAKE, sadly... this is what we are also doing. cake is finally just a option. you can select multiple schedulers at the gui. including codel. fq_codel, fq_codel_fast, cake , pie etc. > > -Toke >