From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-pj1-x1035.google.com (mail-pj1-x1035.google.com [IPv6:2607:f8b0:4864:20::1035]) (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 DF7523B2A4 for ; Sat, 12 Nov 2022 12:19:18 -0500 (EST) Received: by mail-pj1-x1035.google.com with SMTP id e7-20020a17090a77c700b00216928a3917so10268910pjs.4 for ; Sat, 12 Nov 2022 09:19:18 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20210112; h=cc:subject:message-id:date:from:in-reply-to:references:mime-version :from:to:cc:subject:date:message-id:reply-to; bh=hZcJ+tN2991PynX+VDJE+ThooFv3BHFCeBKwUSi+VFM=; b=YPW3sSIg0OTIdQlNkQqNnuTrAsng6rLC9AePlswmNtGLSSE/cdp7Vfht7k+RaiRaHe mYq8f675EzRjjed3HAi8LcB+9zoLlhJlHJDvj6PX5YHo6jDRvEYpbblqpu+QFJldTHHf DM+7PYkesOGm9If15gm0paPX9ECV1aLbwgs9MRKtykoISDHCntv19MI3/mqpmq4lrvdu m5l4JJ7cIU/J/4vpepLFILogwuAvN8aCeDE4lQBOgXzVdazlyrFSY6iuzHEa10YZ9XUH 2K2pEMeywQFgY3y8fH6r+3ybY4IajYnBlm6DvtV/zD8pBwcmbaenf+ypwvaDhjoD2Bcz 4iuw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=cc:subject:message-id:date:from:in-reply-to:references:mime-version :x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=hZcJ+tN2991PynX+VDJE+ThooFv3BHFCeBKwUSi+VFM=; b=F5mrfuWrkfD6KMxpGXgjXsYxezQGZNs1aEtL7x1s+ZsJlYABdUWghr/DRZNWsKNhu2 7vIfyVHf/mWSIr5osxcPQzGbD6d874woGLNtx4XCcbGxLoIvS4WDXXMM+0ujnvd8QjfM gyc74otMRifUdwlM0qvWPMuAuqk13uaE692OEiUulLN8iH1J30qCDLJbZ54Jmqr0fHHk 2XT1eFJSXm3cFt8qlUaNwtfHeqjIxghfDiEu9j6Dh1laj/UhgkdsEBuV66XdLihfcR7w JmagngngEtlKHABCKUAbiQLd4xtqYP0igHojOIpC3AruXDxmQ3Lv56xP5sTKE0Ch2ujl sWIA== X-Gm-Message-State: ANoB5pnukXrdqhwquZzoIVpjRY6tj0Qmv5wLRLtaJ9t3EMZO0VEnrXMy knGdEBqKNNE9crFqzDJZpD7dZtOEqvG1qDwTPsWfXhaZ6yM= X-Google-Smtp-Source: AA0mqf64e4GQJZVzlLmGRna9a9zcfAuSFq8x84fYc+7DlN3IknjYNF30zZ6AcbAbC6BYN0Kwi5jQLzsZP2nZwcYRkWc= X-Received: by 2002:a17:90b:254b:b0:205:a550:e529 with SMTP id nw11-20020a17090b254b00b00205a550e529mr7230379pjb.25.1668273557331; Sat, 12 Nov 2022 09:19:17 -0800 (PST) MIME-Version: 1.0 References: <938D9D45-DADA-4291-BD8A-84E4257CEE49@apple.com> In-Reply-To: From: Herbert Wolverson Date: Sat, 12 Nov 2022 11:19:07 -0600 Message-ID: Cc: libreqos Content-Type: multipart/alternative; boundary="000000000000c5e2cc05ed4934a6" Subject: Re: [LibreQoS] Fwd: A quick report from the WISPA conference X-BeenThere: libreqos@lists.bufferbloat.net X-Mailman-Version: 2.1.20 Precedence: list List-Id: Many ISPs need the kinds of quality shaping cake can do List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 12 Nov 2022 17:19:19 -0000 --000000000000c5e2cc05ed4934a6 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Net neutrality is complex (I'm generally in favor of the concept). There's nothing stopping an ISP from having rules that video streaming be limited - but ALL video streaming has to be equally limited. So ComCast could use Sandvine, but only without any vendor-specific limitations. Same for Cambium. Limit streams that look like video, and you're fine. With that said, it runs into something I say a lot: don't penalize your customers for using your product. If you're buying a 100mbit/s pipe, hooking it to your fancy 4k streaming system - it's entirely reasonable to expect sharp 4k video. In fact, your shaping emphasis should be on keeping their video nice and watchable even though someone just decided to download the newest Call of Duty, update their xbox, and update the firmware on their smart fridge. Likewise, if someone really wants the smallest 5 mbit connection you once offered - because they only ever use it to check email (we actually had a customer complain when we offered them a free upgrade from a 5mbit/s plan, years ago!) - you want to do your best to make it a usable tiny plan for them. Fair queuing (along with a bit of user education; we eventually convinced the customer that "no change in fees" actually meant not charging them more money) helps with all of that. It can win customers, too. We have a couple of business customers who switched us from their backup to their primary because "our service felt snappier". We had a business go with us because their 24x7 lobby streaming video looked better on our (smaller) connection, and a couple of fraternities who likewise decided to use us because their overkill 4k video setup looked better (even with the 20+ Xboxes we could see on their network). QUIC makes me chuckle, because it's *exactly* what we were doing in gamedev land in the late 90s to make deathmatches run smoothly. (Seriously, QuakeWorld and the original Unreal Tournament had the most amazing network code; they basically implemented what is now known as "reliable UDP" to incredible results). With that said, I think we've got a remarkable amount of wiggle-room to make things better (and it's not at all shoddy right now!). XDP is amazing as-is, and is improving fast. (I'm currently reading https://github.com/ilejn/xdpbridge - an older project that doesn't seem to have gained much traction, but I'm wondering if it couldn't provide some seriously fast bridge offloading; I'll have to setup a better test environment to find out) On Sat, Nov 12, 2022 at 10:09 AM Robert Chac=C3=B3n via LibreQoS < libreqos@lists.bufferbloat.net> wrote: > Thank you, Dan. > I completely agree with you 100% there. It's a risky way to save a tiny > amount of money on bandwidth costs. > Comcast paid out $16 million in a settlement > > over using Sandvine's DPI in this way. > Since then, Comcast has switched from DPI to fair queueing with DOCSIS-pi= e > > . > > California's SB-822 > > will be used as a template by many other states going forward on Net > Neutrality. > > SB-822 even clarified that "Reasonable Network Management" needs to be "a= s > application-agnostic as possible". > These DPI approaches like Cambium's will not hold up in court the way Fai= r > Queuing (Preseem, LibreQoS) based solutions will. > We can point to years of public research and RFCs when defending a Fair > Queuing QoE approach. > They can't so easily defend using a product that advertised its big "scal= e > down 4k to save money" knob or "accelerate their speeds only when they ru= n > speed tests" button. > It's concerning to me that these QoE vendors are not disclosing "hey, > don't use this in NN states like California please". > The WISP industry will be hurt long-term by these products with consumers > viewing us as an inferior technology that limits their usage in intrusive > ways. > Like you said, let's let them use their plan however they want. It doesn'= t > hurt us. But things like capping end-users at HD will hurt us long-term i= n > lost sales. > > Regarding TCP acceleration, I'm leaning toward the argument made by Dan > from Preseem that QUIC adoption i= s > growing pretty quickly and we may not want to mess with TCP at the expens= e > of UDP given the rise of QUIC. > Cloudlfare says HTTP/3 and QUIC are already at 28% and growing > . > One point he brought up is that if we use TCP acceleration, and a > middlebox has to be rebooted or something - all client TCP sessions will > break when OSPF switches paths around the box. > > On Sat, Nov 12, 2022 at 8:36 AM dan via LibreQoS < > libreqos@lists.bufferbloat.net> wrote: > >> I think there's a thin grey line between what other vendors (paraqum, >> bequant/cambium) and libreQoS is trying to do. The first two are REALLY >> pushing this selecting narrowing of traffic and they'll go straight afte= r >> streaming for their demos. Most people that drop in the cambium QoE >> appliance turn the knobs on Netflix down and then praise the 40% savings= on >> their network. Same with PQ. I think this is fundamentally wrong. Th= is >> is the entire reason Net Neutrality keeps popping up. Why does the ISP = get >> to say "You only need 25Mbps for your netflix". That's double wrong if >> they offer their own video services. It's not about improving the >> customer's experience, it's about spending less on upstream bandwidth at >> the customer's expense. As soon as they say 'pays for itself in bandwid= th >> savings' you know what the product is primarily for. >> >> This is why I like the preseem and libreqos model of just using a 'stock= ' >> cake and just finessing the data coming in to keep the pipe from cloggin= g. >> Who cares if someone pulls 90M bursts of apple TV+ on their 100M pipe. >> Only dial that back so other requests they make get through cleanly. >> >> preseem is going more towards wireless monitoring etc and kinda letting >> shaping be a secondary.... if they were doing more libreqos like stuff I= 'd >> just stick with preseem. >> >> libreQoS may end up being the best and most net neutral product you can >> get. The only current disadvantage (ignoring DPI) is lack of TCP >> accelleration that PQ and Bequant/cambium have. >> >> I should add that 3 states already have laws on the books that straight >> up make PQ and Cambium's solution illegal because they target certain >> services for throttling. My state (Montana) has the same rule except on= ly >> when working with a government contract. Libreqos is actually compliant >> (as is preseem). >> >> On Sat, Nov 12, 2022 at 8:11 AM Dave Taht via LibreQoS < >> libreqos@lists.bufferbloat.net> wrote: >> >>> this report predates the libreqos list... >>> >>> ---------- Forwarded message --------- >>> From: Dave Taht >>> Date: Mon, Oct 17, 2022 at 8:15 PM >>> Subject: A quick report from the WISPA conference >>> To: Sina Khanifar >>> Cc: Cake List , Make-Wifi-fast >>> , Rpm >>> , Stuart Cheshire , >>> bloat >>> >>> >>> On Mon, Oct 17, 2022 at 7:51 PM Sina Khanifar wrote= : >>> > >>> > Positive or negative, I can claim a bit of credit for this video :). >>> We've been working with LTT on a few projects and we pitched them on do= ing >>> something around bufferbloat. We've seen more traffic to our Waveforn t= est >>> than ever before, which has been fun! >>> >>> Thank you. Great job with that video! And waveform has become the goto >>> site for many now. >>> >>> I can't help but wonder tho... are you collecting any statistics, over >>> time, as to how much better the problem is getting? >>> >>> And any chance they could do something similar explaining wifi? >>> >>> ... >>> >>> I was just at WISPA conference week before last. Preseem's booth >>> (fq_codel) was always packed. Vilo living had put cake in their wifi 6 >>> product. A >>> keynote speaker had deployed it and talked about it with waveform >>> results on the big screen (2k people there). A large wireless vendor >>> demo'd privately to me their flent results before/after cake on their >>> next-gen radios... and people dissed tarana without me prompting for >>> their bad bufferbloat... and the best thing of all that happened to me >>> was... besides getting a hug from a young lady (megan) who'd salvaged >>> her schooling in alaska using sqm - I walked up to the paraqum booth >>> (another large QoE middlebox maker centered more in india) and asked. >>> >>> "So... do y'all have fq_codel yet?" >>> >>> And they smiled and said: "No, we have something better... we've got >>> cake." >>> >>> "Cake? What's that?" - I said, innocently. >>> >>> They then stepped me through their 200Gbps (!!) product, which uses a >>> bunch of offloads, and can track rtt down to a ms with the intel >>> ethernet card they were using. They'd modifed cake to provide 16 (?) >>> levels of service, and were running under dpdk (I am not sure if cake >>> was). It was a great, convincing pitch... >>> >>> ... then I told 'em who I was. There's a video of the in-both concert >>> after. >>> >>> ... >>> >>> The downside to me (and the subject of my talk) was that in nearly >>> every person I talked to, fq_codel was viewed as a means to better >>> subscriber bandwidth plan enforcement (which is admittedly the market >>> that preseem pioneered) and it was not understood that I'd got >>> involved in this whole thing because I'd wanted an algorithm to deal >>> with "rain fade", running directly on the radios. People wanted to use >>> the statistics on the radios to drive the plan enforcement better >>> (which is an ok approach, I guess), and for 10+ I'd been whinging >>> about the... physics. >>> >>> So I ranted about rfc7567 a lot and begged people now putting routerOS >>> 7.2 and later out there (mikrotik is huge in this market), to kill >>> their fifos and sfqs at the native rates of the interfaces... and >>> watch their network improve that way also. >>> >>> I think one more wispa conference will be a clean sweep of everyone in >>> the fixed wireless market to not only adopt these algorithms for plan >>> enforcement, but even more directly on the radios and more CPE. >>> >>> I also picked up enough consulting business to keep me busy the rest >>> of this year, and possibly more than I can handle (anybody looking?) >>> >>> I wonder what will happen at a fiber conference? >>> >>> > On Mon, Oct 17, 2022 at 7:45 PM Dave Taht via Bloat < >>> bloat@lists.bufferbloat.net> wrote: >>> >> >>> >> On Mon, Oct 17, 2022 at 5:02 PM Stuart Cheshire >>> wrote: >>> >> > >>> >> > On 9 Oct 2022, at 06:14, Dave Taht via Make-wifi-fast < >>> make-wifi-fast@lists.bufferbloat.net> wrote: >>> >> > >>> >> > > This was so massively well done, I cried. Does anyone know how t= o >>> get in touch with the ifxit folk? >>> >> > > >>> >> > > https://www.youtube.com/watch?v=3DUICh3ScfNWI >>> >> > >>> >> > I=E2=80=99m surprised that you liked this video. It seems to me th= at it >>> repeats all the standard misinformation. The analogy they use is the >>> standard terrible example of waiting in a long line at a grocery store,= and >>> the =E2=80=9Csolution=E2=80=9D is letting certain traffic =E2=80=9Cjump= the line, angering everyone >>> behind them=E2=80=9D. >>> >> >>> >> Accuracy be damned. The analogy to common experience resonates more. >>> >> >>> >> > >>> >> > Some quotes from the video: >>> >> > >>> >> > > it would be so much more efficient for them to let you skip the >>> line and just check out, especially since you=E2=80=99re in a hurry, bu= t they=E2=80=99re >>> rudely refusing >>> >> >>> >> I think the person with the cheetos pulling out a gun and shooting >>> >> everyone in front of him (AQM) would not go down well. >>> >> >>> >> > > to go back to our grocery store analogy this would be like if a >>> worker saw you standing at the back ... and either let you skip to the >>> front of the line or opens up an express lane just for you >>> >> >>> >> Actually that analogy is fairly close to fair queuing. The multiple >>> >> checker analogy is one of the most common analogies in queue theory >>> >> itself. >>> >> >>> >> > >>> >> > The video describes the problem of bufferbloat, and then describes >>> the same failed solution that hasn=E2=80=99t worked for the last three = decades. >>> >> >>> >> Hmm? It establishes the scenario, explains the problem *quickly*, >>> >> disses gamer routers for not getting it right.. *points to an >>> >> accurate test*, and then to the ideas and products that *actually >>> >> work* with "smart queueing", with a screenshot of the most common >>> >> (eero's optimize for gaming and videoconferencing), and fq_codel and >>> >> cake *by name*, and points folk at the best known solution available= , >>> >> openwrt. >>> >> >>> >> Bing, baddabang, boom. Also the comments were revealing. A goodly >>> >> percentage already knew the problem, more than a few were inspired t= o >>> >> take the test, >>> >> there was a whole bunch of "Aha!" success stories and 360k views, >>> >> which is more people than we've ever been able to reach in for >>> >> example, a nanog conference. >>> >> >>> >> I loved that folk taking the test actually had quite a few A results= , >>> >> without having had to do anything. At least some ISPs are getting it >>> >> more right now! >>> >> >>> >> At this point I think gamers in particular know what "brands" we've >>> >> tried to establish - "Smart queues", "SQM", "OpenWrt", fq_codel and >>> >> now "cake" are "good" things to have, and are stimulating demand by >>> >> asking for them, It's certainly working out better and better for >>> >> evenroute, firewalla, ubnt and others, and I saw an uptick in >>> >> questions about this on various user forums. >>> >> >>> >> I even like that there's a backlash now of people saying "fixing >>> >> bufferbloat doesn't solve everything" - >>> >> >>> >> > Describing the obvious simple-minded (wrong) solution that any >>> normal person would think of based on their personal human experience >>> waiting in grocery stores and airports, is not describing the solution = to >>> bufferbloat. The solution to bufferbloat is not that if you are privile= ged >>> then you get to =E2=80=9Cskip to the front of the line=E2=80=9D. The so= lution to >>> bufferbloat is that there is no line! >>> >> >>> >> I like the idea of a guru floating above a grocery cart with a bette= r >>> >> string of explanations, explaining >>> >> >>> >> - "no, grasshopper, the solution to bufferbloat is no line... at >>> all". >>> >> >>> >> > >>> >> > With grocery stores and airports people=E2=80=99s arrivals are ind= ependent >>> and not controlled. There is no way for a grocery store or airport to >>> generate backpressure to tell people to wait at home when a queue begin= s to >>> form. The key to solving bufferbloat is generating timely backpressure = to >>> prevent the queue forming in the first place, not accepting a huge queu= e >>> and then deciding who deserves special treatment to get better service = than >>> all the other peons who still have to wait in a long queue, just like >>> before. >>> >> >>> >> I am not huge on the word "backpressure" here. Needs to signal the >>> >> other side to slow down, is more accurate. So might say timely >>> >> signalling rather than timely backpressure? >>> >> >>> >> Other feedback I got was that the video was too smarmy (I agree), >>> >> different audiences than gamers need different forms of outreach... >>> >> >>> >> but to me, winning the gamers has always been one of the most >>> >> important things, as they make a lot of buying decisions, and they >>> >> benefit the most for >>> >> fq and packet prioritization as we do today in gamer routers and in >>> >> cake + qosify. >>> >> >>> >> maybe that gets in the way of more serious markets. Certainly I woul= d >>> >> like another video explaining what goes wrong with videoconferencing= . >>> >> >>> >> >>> >> >>> >> >>> >> >>> >> >>> >> > >>> >> > Stuart Cheshire >>> >> > >>> >> >>> >> >>> >> -- >>> >> This song goes out to all the folk that thought Stadia would work: >>> >> >>> https://www.linkedin.com/posts/dtaht_the-mushroom-song-activity-6981366= 665607352320-FXtz >>> >> Dave T=C3=A4ht CEO, TekLibre, LLC >>> >> _______________________________________________ >>> >> Bloat mailing list >>> >> Bloat@lists.bufferbloat.net >>> >> https://lists.bufferbloat.net/listinfo/bloat >>> >>> >>> >>> -- >>> This song goes out to all the folk that thought Stadia would work: >>> >>> https://www.linkedin.com/posts/dtaht_the-mushroom-song-activity-6981366= 665607352320-FXtz >>> Dave T=C3=A4ht CEO, TekLibre, LLC >>> >>> >>> -- >>> This song goes out to all the folk that thought Stadia would work: >>> >>> https://www.linkedin.com/posts/dtaht_the-mushroom-song-activity-6981366= 665607352320-FXtz >>> Dave T=C3=A4ht CEO, TekLibre, LLC >>> _______________________________________________ >>> LibreQoS mailing list >>> LibreQoS@lists.bufferbloat.net >>> https://lists.bufferbloat.net/listinfo/libreqos >>> >> _______________________________________________ >> LibreQoS mailing list >> LibreQoS@lists.bufferbloat.net >> https://lists.bufferbloat.net/listinfo/libreqos >> > > > -- > Robert Chac=C3=B3n > CEO | JackRabbit Wireless LLC > Dev | LibreQoS.io > > _______________________________________________ > LibreQoS mailing list > LibreQoS@lists.bufferbloat.net > https://lists.bufferbloat.net/listinfo/libreqos > --000000000000c5e2cc05ed4934a6 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable
Net neutrality is complex (I'm generally in favor= of the concept). There's nothing stopping an ISP from having rules tha= t video streaming be limited - but ALL video streaming has to be equally li= mited. So ComCast could use Sandvine, but only without any vendor-specific = limitations. Same for Cambium. Limit streams that look like video, and you&= #39;re fine. With that said, it runs into something I say a lot: don't = penalize your customers for using your product. If you're buying a 100m= bit/s pipe, hooking it to your fancy 4k streaming system - it's entirel= y reasonable to expect sharp 4k video. In fact, your shaping emphasis shoul= d be on keeping their video nice and watchable even though someone just dec= ided to download the newest Call of Duty, update their xbox, and update the= firmware on their smart fridge. Likewise, if someone really wants the smal= lest 5 mbit connection you once offered - because they only ever use it to = check email (we actually had a customer complain when we offered them a fre= e upgrade from a 5mbit/s plan, years ago!) - you want to do your best to ma= ke it a usable tiny plan for them. Fair queuing (along with a bit of user e= ducation; we eventually convinced the customer that "no change in fees= " actually meant not charging them more money) helps with all of that.=

It can win customers, too. We have a couple = of business customers who switched us from their backup to their primary be= cause "our service felt snappier". We had a business go with us b= ecause their 24x7 lobby streaming video looked better on our (smaller) conn= ection, and a couple of fraternities who likewise decided to use us because= their overkill 4k video setup looked better (even with the 20+ Xboxes we c= ould see on their network).

QUIC makes me chuckle,= because it's exactly what we were doing in gamedev land in the = late 90s to make deathmatches run smoothly. (Seriously, QuakeWorld and the = original Unreal Tournament had the most amazing network code; they basicall= y implemented what is now known as "reliable UDP" to incredible r= esults).

With that said, I think we've got a r= emarkable amount of wiggle-room to make things better (and it's not at = all shoddy right now!). XDP is amazing as-is, and is improving fast. (I'= ;m currently reading https:/= /github.com/ilejn/xdpbridge - an older project that doesn't seem to= have gained much traction, but I'm wondering if it couldn't provid= e some seriously fast bridge offloading; I'll have to setup a better te= st environment to find out)

<= div dir=3D"ltr" class=3D"gmail_attr">On Sat, Nov 12, 2022 at 10:09 AM Rober= t Chac=C3=B3n via LibreQoS <libreqos@lists.bufferbloat.net> wrote:
Thank you, Dan.
I completely agree=C2=A0with you 100% there. It's a risky way = to save a tiny amount of money on bandwidth costs.
Comcast paid o= ut $16 million in a settlement over using Sandvine's DPI i= n this way.
Since then, Comcast has switched from DPI to fair que= ueing with DOCSIS-pie.

California's SB-822 will be used = as a template by many other states going forward on Net Neutrality.<= /a>
SB-822 even clarified that "Reasonable Network Management" needs to be = "as=20 application-agnostic as possible".
These DPI approaches like Cambium's will not hold up in = court the way Fair Queuing (Preseem, LibreQoS) based=20 solutions will.
We can po= int to years of public research and RFCs when defending a Fair Queuing QoE approach.
They can't so easily defend using a product that advertised its big "scale down 4k to save money" kn= ob or=20 "accelerate their speeds only when they run speed tests" button.<= /div>
It's concerning to me that these Q= oE vendors are not disclosing "hey, don't use this in NN states li= ke California please".
The WISP i= ndustry will be hurt long-term by these products with consumers viewing us = as an inferior technology that limits their usage in intrusive ways.
<= div style=3D"text-align:start">Like you said, let's let them use their = plan however they want. It doesn't hurt us. But things like capping end= -users at HD will hurt us long-term in lost sales.

Reg= arding TCP acceleration, I'm leaning toward the argument made by Dan from Pres= eem that QUIC adoption is growing pretty quickly and we may not want to= mess with TCP at the expense of UDP given the rise of QUIC.
Cloudlfare says HTTP/3 and QUIC are already at 28% and growing.
O= ne point he brought up is that if we use TCP acceleration, and a middlebox = has to be rebooted or something - all client TCP sessions will break when O= SPF switches paths around the box.

On Sat, Nov 12, = 2022 at 8:36 AM dan via LibreQoS <libreqos@lists.bufferbloat.net> wrote:=
I think there's a thin grey line between what other vendors (paraqum,= bequant/cambium) and=C2=A0libreQoS is trying to do.=C2=A0 The first two ar= e REALLY pushing this selecting narrowing of traffic and they'll go str= aight after streaming for their demos.=C2=A0 Most people that drop in the c= ambium QoE appliance turn the knobs on Netflix down and then praise the 40%= savings on their network.=C2=A0 Same with PQ.=C2=A0 =C2=A0I think this is = fundamentally wrong.=C2=A0 This is the entire reason Net Neutrality keeps p= opping up.=C2=A0 Why does the ISP get to say "You only need 25Mbps for= your netflix".=C2=A0 That's double wrong if they offer their own = video services.=C2=A0 It's not about improving the customer's exper= ience, it's about spending less on upstream bandwidth at the customer&#= 39;s expense.=C2=A0 As soon as they say 'pays for itself in bandwidth s= avings' you know what the product is primarily for.

This is why = I like the preseem and libreqos model of just using a 'stock' cake = and just finessing the data coming in to keep the pipe from clogging.=C2=A0= Who cares if someone pulls 90M bursts of apple TV+=C2=A0on their 100M pipe= .=C2=A0 Only dial that back so other requests they make get through cleanly= .

preseem is going more towards wireless monitoring etc and kinda le= tting shaping be a secondary.... if they were doing more libreqos like stuf= f I'd just stick with preseem.

libreQoS=C2=A0may end up being th= e best and most net neutral product you can get.=C2=A0 The only current dis= advantage (ignoring DPI) is lack of TCP accelleration=C2=A0that PQ and Bequ= ant/cambium have.

I should add that 3 states already have laws on th= e books that straight up make PQ and Cambium's solution illegal because= they target certain services for throttling.=C2=A0 My state (Montana) has = the same rule except only when working with a government contract.=C2=A0 Li= breqos=C2=A0is actually compliant (as is preseem).

On Sat, Nov 12, 2022 at 8= :11 AM Dave Taht via LibreQoS <libreqos@lists.bufferbloat.net> wrote:
this report predat= es the libreqos list...

---------- Forwarded message ---------
From: Dave Taht <dave.taht@gmail.com>
Date: Mon, Oct 17, 2022 at 8:15 PM
Subject: A quick report from the WISPA conference
To: Sina Khanifar <sina@waveform.com>
Cc: Cake List <cake@lists.bufferbloat.net>, Make-Wifi-fast
<make-wifi-fast@lists.bufferbloat.net>, Rpm
<rpm@list= s.bufferbloat.net>, Stuart Cheshire <cheshire@apple.com>,
bloat <= bloat@lists.bufferbloat.net>


On Mon, Oct 17, 2022 at 7:51 PM Sina Khanifar <sina@waveform.com> wrote:
>
> Positive or negative, I can claim a bit of credit for this video :). W= e've been working with LTT on a few projects and we pitched them on doi= ng something around bufferbloat. We've seen more traffic to our Wavefor= n test than ever before, which has been fun!

Thank you. Great job with that video! And waveform has become the goto
site for many now.

I can't help but wonder tho... are you collecting any statistics, over<= br> time, as to how much better the problem is getting?

And any chance they could do something similar explaining wifi?

...

I was just at WISPA conference week before last. Preseem's booth
(fq_codel) was always packed. Vilo living had put cake in their wifi 6
product. A
keynote speaker had deployed it and talked about it with waveform
results on the big screen (2k people there). A large wireless vendor
demo'd privately to me their flent results before/after cake on their next-gen radios... and people dissed tarana without me prompting for
their bad bufferbloat... and the best thing of all that happened to me
was... besides getting a hug from a young lady (megan) who'd salvaged her schooling in alaska using sqm - I walked up to the paraqum booth
(another large QoE middlebox maker centered more in india) and asked.

"So... do y'all have fq_codel yet?"

And they smiled and said: "No, we have something better... we've g= ot cake."

"Cake? What's that?" - I said, innocently.

They then stepped me through their 200Gbps (!!) product, which uses a
bunch of offloads, and can track rtt down to a ms with the intel
ethernet card they were using. They'd modifed cake to provide 16 (?) levels of service, and were running under dpdk (I am not sure if cake
was). It was a great, convincing pitch...

... then I told 'em who I was. There's a video of the in-both conce= rt after.

...

The downside to me (and the subject of my talk) was that in nearly
every person I talked to, fq_codel was viewed as a means to better
subscriber bandwidth plan enforcement (which is admittedly the market
that preseem pioneered) and it was not understood that I'd got
involved in this whole thing because I'd wanted an algorithm to deal with "rain fade", running directly on the radios. People wanted t= o use
the statistics on the radios to drive the plan enforcement better
(which is an ok approach, I guess), and for 10+ I'd been whinging
about the... physics.

So I ranted about rfc7567 a lot and begged people now putting routerOS
7.2 and later out there (mikrotik is huge in this market), to kill
their fifos and sfqs at the native rates of the interfaces... and
watch their network improve that way also.

I think one more wispa conference will be a clean sweep of everyone in
the fixed wireless market to not only adopt these algorithms for plan
enforcement, but even more directly on the radios and more CPE.

I also picked up enough consulting business to keep me busy the rest
of this year, and possibly more than I can handle (anybody looking?)

I wonder what will happen at a fiber conference?

> On Mon, Oct 17, 2022 at 7:45 PM Dave Taht via Bloat <bloat@lists.bufferbloat.= net> wrote:
>>
>> On Mon, Oct 17, 2022 at 5:02 PM Stuart Cheshire <cheshire@apple.com> wrote:=
>> >
>> > On 9 Oct 2022, at 06:14, Dave Taht via Make-wifi-fast <make= -wifi-fast@lists.bufferbloat.net> wrote:
>> >
>> > > This was so massively well done, I cried. Does anyone kn= ow how to get in touch with the ifxit folk?
>> > >
>> > > https://www.youtube.com/watch?v=3DUI= Ch3ScfNWI
>> >
>> > I=E2=80=99m surprised that you liked this video. It seems to = me that it repeats all the standard misinformation. The analogy they use is= the standard terrible example of waiting in a long line at a grocery store= , and the =E2=80=9Csolution=E2=80=9D is letting certain traffic =E2=80=9Cju= mp the line, angering everyone behind them=E2=80=9D.
>>
>> Accuracy be damned. The analogy to common experience resonates mor= e.
>>
>> >
>> > Some quotes from the video:
>> >
>> > > it would be so much more efficient for them to let you s= kip the line and just check out, especially since you=E2=80=99re in a hurry= , but they=E2=80=99re rudely refusing
>>
>> I think the person with the cheetos pulling out a gun and shooting=
>> everyone in front of him (AQM) would not go down well.
>>
>> > > to go back to our grocery store analogy this would be li= ke if a worker saw you standing at the back ... and either let you skip to = the front of the line or opens up an express lane just for you
>>
>> Actually that analogy is fairly close to fair queuing. The multipl= e
>> checker analogy is one of the most common analogies in queue theor= y
>> itself.
>>
>> >
>> > The video describes the problem of bufferbloat, and then desc= ribes the same failed solution that hasn=E2=80=99t worked for the last thre= e decades.
>>
>> Hmm? It establishes the scenario, explains the problem *quickly*,<= br> >> disses gamer routers for not getting it right..=C2=A0 *points to a= n
>> accurate test*, and then to the ideas and products that *actually<= br> >> work* with "smart queueing", with a screenshot of the mo= st common
>> (eero's optimize for gaming and videoconferencing), and fq_cod= el and
>> cake *by name*, and points folk at the best known solution availab= le,
>> openwrt.
>>
>> Bing, baddabang, boom. Also the comments were revealing. A goodly<= br> >> percentage already knew the problem, more than a few were inspired= to
>> take the test,
>> there was a whole bunch of "Aha!" success stories and 36= 0k views,
>> which is more people than we've ever been able to reach in for=
>> example, a nanog conference.
>>
>> I loved that folk taking the test actually had quite a few A resul= ts,
>> without having had to do anything. At least some ISPs are getting = it
>> more right now!
>>
>> At this point I think gamers in particular know what "brands&= quot; we've
>> tried to establish - "Smart queues", "SQM", &q= uot;OpenWrt", fq_codel and
>> now "cake" are "good" things to have, and are = stimulating demand by
>> asking for them,=C2=A0 =C2=A0It's certainly working out better= and better for
>> evenroute, firewalla, ubnt and others, and I saw an uptick in
>> questions about this on various user forums.
>>
>> I even like that there's a backlash now of people saying "= ;fixing
>> bufferbloat doesn't solve everything" -
>>
>> >=C2=A0 Describing the obvious simple-minded (wrong) solution t= hat any normal person would think of based on their personal human experien= ce waiting in grocery stores and airports, is not describing the solution t= o bufferbloat. The solution to bufferbloat is not that if you are privilege= d then you get to =E2=80=9Cskip to the front of the line=E2=80=9D. The solu= tion to bufferbloat is that there is no line!
>>
>> I like the idea of a guru floating above a grocery cart with a bet= ter
>> string of explanations, explaining
>>
>>=C2=A0 =C2=A0 - "no, grasshopper, the solution to bufferbloat = is no line... at all".
>>
>> >
>> > With grocery stores and airports people=E2=80=99s arrivals ar= e independent and not controlled. There is no way for a grocery store or ai= rport to generate backpressure to tell people to wait at home when a queue = begins to form. The key to solving bufferbloat is generating timely backpre= ssure to prevent the queue forming in the first place, not accepting a huge= queue and then deciding who deserves special treatment to get better servi= ce than all the other peons who still have to wait in a long queue, just li= ke before.
>>
>> I am not huge on the word "backpressure" here. Needs to = signal the
>> other side to slow down, is more accurate. So might say timely
>> signalling rather than timely backpressure?
>>
>> Other feedback I got=C2=A0 was that the video was too smarmy (I ag= ree),
>> different audiences than gamers need different forms of outreach..= .
>>
>> but to me, winning the gamers has always been one of the most
>> important things, as they make a lot of buying decisions, and they=
>> benefit the most for
>> fq and packet prioritization as we do today in gamer routers and i= n
>> cake + qosify.
>>
>> maybe that gets in the way of more serious markets. Certainly I wo= uld
>> like another video explaining what goes wrong with videoconferenci= ng.
>>
>>
>>
>>
>>
>>
>> >
>> > Stuart Cheshire
>> >
>>
>>
>> --
>> This song goes out to all the folk that thought Stadia would work:=
>> htt= ps://www.linkedin.com/posts/dtaht_the-mushroom-song-activity-69813666656073= 52320-FXtz
>> Dave T=C3=A4ht CEO, TekLibre, LLC
>> _______________________________________________
>> Bloat mailing list
>> B= loat@lists.bufferbloat.net
>> https://lists.bufferbloat.net/listinfo/bloat



--
This song goes out to all the folk that thought Stadia would work:
https://www.= linkedin.com/posts/dtaht_the-mushroom-song-activity-6981366665607352320-FXt= z
Dave T=C3=A4ht CEO, TekLibre, LLC


--
This song goes out to all the folk that thought Stadia would work:
https://www.= linkedin.com/posts/dtaht_the-mushroom-song-activity-6981366665607352320-FXt= z
Dave T=C3=A4ht CEO, TekLibre, LLC
_______________________________________________
LibreQoS mailing list
LibreQo= S@lists.bufferbloat.net
https://lists.bufferbloat.net/listinfo/libreqos
_______________________________________________
LibreQoS mailing list
LibreQo= S@lists.bufferbloat.net
https://lists.bufferbloat.net/listinfo/libreqos


--
Robert Chac=C3=B3n

<= /div>
_______________________________________________
LibreQoS mailing list
LibreQo= S@lists.bufferbloat.net
https://lists.bufferbloat.net/listinfo/libreqos
--000000000000c5e2cc05ed4934a6--