From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-pf1-x432.google.com (mail-pf1-x432.google.com [IPv6:2607:f8b0:4864:20::432]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by lists.bufferbloat.net (Postfix) with ESMTPS id 68B653B29D for ; Wed, 19 Oct 2022 19:36:18 -0400 (EDT) Received: by mail-pf1-x432.google.com with SMTP id y1so18687612pfr.3 for ; Wed, 19 Oct 2022 16:36:18 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=networkplumber-org.20210112.gappssmtp.com; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:subject:cc:to:from:date:from:to:cc:subject:date :message-id:reply-to; bh=HZpHDo8oT6OkNR+ftLPy11oEvPqqf0QKc+lf3yYatgc=; b=cJDXmFtYAHlWrxIsVY3Wb95AVtjQtxQQqmuwGtN3BsFYo6UNBtXo0vkd5jRZNtxRHm kszuhK2X+CKBLi8st6ZjLCpV6V1bY9r23XL6UI6q10zr+0wr6WKEGqxPNtDFxns9HL4k sxmOVab8oBaFj+n8qve5vB0awjS4A9zp2RgUrShHROtWcqJdvQaKLVzQDqeFyI3anvk3 c0XLnm+6/YhD3mfjBUFroWTf7O2k5leEpcNHs1wB9mNHy8LnJsErL9pItD23CXaoe3UP TCRiGLABkSy1Fyz/v2esDl51LmXbOBdvP96TwxB9ZraOAhjEUytVd2Rx8reZm0mxiCFL 5F2w== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:subject:cc:to:from:date:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=HZpHDo8oT6OkNR+ftLPy11oEvPqqf0QKc+lf3yYatgc=; b=De3tTtSVvxnHcXVD3se/RH+76kJlHg90bjqyWtQ9nSi9EEqp5V4Jt4PQrV+URE9KQT Ab7IrOrJttiWsw4tswKb48D2BFvEtf04qk8gRtjyQtOnjsCfDiRuVM9l7AmFe2CsFC0q 3idsZrxYvkdywDTy6x3igL3XTicGTx8YRm65KTRWm4LNOQ8XZ/jcA7Y2w4pSEwKyiJ67 kqASO9NO1aUO+ziG55YVoL+WWxfRa15rv0BqjgnvgxrjIy7V3R2+4o2cAshN/uxGzS+i YSRGMX4ZhBBOA0vhm4NSbeaCuDmP16c+Z9n+JZmaxzw0hVImjNMtiy/ZN26PhJaX6xlN fq8Q== X-Gm-Message-State: ACrzQf15LxoWk0OMy2AaP8sVbMK86rPmoUhfjMOavFS8KW4OJOfPa6Lb y4n8uET7z7CG5Vinux38Q6ZxJKVG1E+E3Q== X-Google-Smtp-Source: AMsMyM4Mz+xtiLq/DB6O/dgJ45WvE9dFPmckrVtawLxYSJatabg8R3NdUOHkj9di9Xa7Lp3PTxsBOw== X-Received: by 2002:a63:1455:0:b0:438:e26b:ab1f with SMTP id 21-20020a631455000000b00438e26bab1fmr9042264pgu.183.1666222577119; Wed, 19 Oct 2022 16:36:17 -0700 (PDT) Received: from hermes.local (204-195-120-218.wavecable.com. [204.195.120.218]) by smtp.gmail.com with ESMTPSA id 21-20020a17090a035500b0020aedddd3cesm503056pjf.5.2022.10.19.16.36.16 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 19 Oct 2022 16:36:16 -0700 (PDT) Date: Wed, 19 Oct 2022 16:36:14 -0700 From: Stephen Hemminger To: David Lang via Bloat Cc: David Lang , Stuart Cheshire , Rpm , Make-Wifi-fast , Cake List Message-ID: <20221019163614.28437804@hermes.local> In-Reply-To: <45so9803-49n7-q176-4or6-o188873243s0@ynat.uz> References: <938D9D45-DADA-4291-BD8A-84E4257CEE49@apple.com> <45so9803-49n7-q176-4or6-o188873243s0@ynat.uz> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Subject: Re: [Bloat] [Make-wifi-fast] The most wonderful video ever about bufferbloat X-BeenThere: bloat@lists.bufferbloat.net X-Mailman-Version: 2.1.20 Precedence: list List-Id: General list for discussing Bufferbloat List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 19 Oct 2022 23:36:18 -0000 On Wed, 19 Oct 2022 14:33:28 -0700 (PDT) David Lang via Bloat wrote: > On Wed, 19 Oct 2022, Stuart Cheshire via Bloat wrote: >=20 > > On Mon, Oct 17, 2022 at 5:02 PM Stuart Cheshire wr= ote: > > =20 > >> Accuracy be damned. The analogy to common experience resonates more. = =20 > > > > I feel it is not an especially profound insight to observe that, =E2=80= =9Cpeople don=E2=80=99t like waiting in line.=E2=80=9D The conclusion, =E2= =80=9Ctherefore privileged people should get to go to the front,=E2=80=9D d= escribes an airport first class checkin counter, Disney Fastpass, and count= less other analogies from everyday life, all of which are the wrong solutio= n for packets in a network. =20 >=20 > the 'privileged go first' is traditional QoS, and it can work to some ext= ent,=20 > but is a nightmare to maintain and gets the wrong result most of the time. A lot of times when this is proposed it has some business/political motivat= ion. It is like "priority boarding" for Global Services customers. Not solving a latency problem, instead making stakeholders happy. > AQM (fw_codel and cake) are more the 'cash only line' and '15 items or le= ss'=20 > line, they speed up the things that can be fast a LOT, while not signific= antly=20 > slowing down the people with a full baskets (but in the process, it short= ens the=20 > lines for those people with full baskets) >=20 > >> I think the person with the cheetos pulling out a gun and shooting eve= ryone in front of him (AQM) would not go down well. =20 > > > > Which is why starting with a bad analogy (people waiting in a grocery s= tore) inevitably leads to bad conclusions. > > > > If we want to struggle to make the grocery store analogy work, perhaps = we show=20 > > people checking some grocery store app on their smartphone before they = leave=20 > > home, and if they see that a long line is beginning to form they wait u= ntil=20 > > later, when the line is shorter. The challenge is not how to deal with = a long=20 > > queue when it=E2=80=99s there, it is how to avoid a long queue in the f= irst place. =20 >=20 > only somewhat, you aren't going to have people deciding not to click on a= link=20 > because the network is busy, and if you did try to go that direction, I w= ould=20 > fight you. the prioritization is happening at a much lower level, which i= s hard=20 > to put into an analogy >=20 > even with the 'slowing' of bulk traffic, no traffic is prevented, it's ju= st that=20 > they aren't allowed to monopolize the links. >=20 > This is where the grocery store analogy is weak, the reality would be mor= e like=20 > 'the cashier will only process 30 items before you have to step aside and= let=20 > someone else in', but since no store operates that way, it would be a bad= =20 > analogy. Grocery store analogies also breakdown because packets are not "precious" it is okay to drop packets. A lot of AQM works by doing "drop early and oft= en" instead of "drop late and collapse". >=20 > >> Actually that analogy is fairly close to fair queuing. The multiple ch= ecker analogy is one of the most common analogies in queue theory itself. = =20 > > > > I disagree. You are describing the =E2=80=9CFQ=E2=80=9D part of FQ_CoDe= l. It=E2=80=99s the =E2=80=9CCoDel=E2=80=9D=20 > > part of FQ_CoDel that solves bufferbloat. FQ has been around for a long= time,=20 > > and at best it partially masked the effects of bufferbloat. Having more= queues=20 > > does not solve bufferbloat. Managing the queue(s) better solves bufferb= loat. > > =20 > >> I like the idea of a guru floating above a grocery cart with a better = string of explanations, explaining > >> > >> - "no, grasshopper, the solution to bufferbloat is no line... at all= ". =20 > > > > That is the kind of thing I had in mind. Or a similar quote from The Ma= trix.=20 > > While everyone is debating ways to live with long queues, the guru asks= , =E2=80=9CWhat=20 > > if there were no queues?=E2=80=9D That is the =E2=80=9Cmind blown=E2=80= =9D realization. =20 >=20 > In a world where there is no universal scheduler (and no universal knowle= ge to=20 > base any scheduling decisions on), and where you are going to have malici= ous=20 > actors trying to get more than their fair share, you can't rely on volunt= ary=20 > actions to eliminate the lines. >=20 > There are data transportation apps that work by starting up a large numbe= r of=20 > connections in parallel for the highest transfer speeds (shortening slow = start,=20 > reducing the impact of lost packets as they only affect one connection, e= tc).=20 > This isn't even malicious actors, but places like Hollywood studios sendi= ng=20 > the raw movie footage around over dedicated leased lines and wanting to g= et=20 > every bps of bandwidth that they are paying for used. >=20 > David Lang