From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from smtp97.iad3a.emailsrvr.com (smtp97.iad3a.emailsrvr.com [173.203.187.97]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by lists.bufferbloat.net (Postfix) with ESMTPS id E86DA3CB44; Mon, 16 Jun 2025 15:03:28 -0400 (EDT) Received: from app35.wa-webapps.iad3a (relay-webapps.rsapps.net [172.27.255.140]) by smtp5.relay.iad3a.emailsrvr.com (SMTP Server) with ESMTP id 5AFC525320; Mon, 16 Jun 2025 15:03:28 -0400 (EDT) Received: from deepplum.com (localhost.localdomain [127.0.0.1]) by app35.wa-webapps.iad3a (Postfix) with ESMTP id 42782A0085; Mon, 16 Jun 2025 15:03:39 -0400 (EDT) Received: by apps.rackspace.com (Authenticated sender: dpreed@deepplum.com, from: dpreed@deepplum.com) with HTTP; Mon, 16 Jun 2025 15:03:39 -0400 (EDT) X-Auth-ID: dpreed@deepplum.com Date: Mon, 16 Jun 2025 15:03:39 -0400 (EDT) From: "David P. Reed" To: "Frantisek Borsik" Cc: "libreqos" , "bloat" , "Dave Taht via Starlink" , "Jeremy Austin via Rpm" , bloat-ietf@lists.bufferbloat.net, "Cake List" , codel@lists.bufferbloat.net, "Herbert Wolverson" , "=?utf-8?Q?Robert_Chac=C3=B3n?=" MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_20250616150339000000_79771" Importance: Normal X-Priority: 3 (Normal) X-Type: html In-Reply-To: References: X-Client-IP: 209.6.168.128 Message-ID: <1750100619.26922718@apps.rackspace.com> X-Mailer: webmail/19.0.28-RC X-Classification-ID: 0e6162e3-8909-4fa0-bf73-43956e47dc48-1-1 Subject: Re: [Starlink] [Cake] Announcing the LibreQoS Bufferbloat Test Platform X-BeenThere: starlink@lists.bufferbloat.net X-Mailman-Version: 2.1.20 Precedence: list List-Id: "Starlink has bufferbloat. Bad." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 16 Jun 2025 19:03:29 -0000 ------=_20250616150339000000_79771 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable =0AI plan to test this out as "being my own ISP" using one of my personal c= loud servers that is in one of Vultur's hosting sites as the server. (and f= rom various mobile devices as I travel around among WiFi sites, if I can).= =0A =0AThe thing I immediately want to know, though, is: has anybody done a= serious security review of the system? I'm a "trust but verify" kind of pe= rsonality. I don't have the energy or time to do the detailed review. I tru= st that the folks who built this are not black hats - you're my "friends" a= fter all - we care about similar things. But I'd like to verify that trust.= =0A =0AAlso, any suggestions people might have with regard to making it wor= k?=0A =0AI'll send feedback as I explore, promise! Great work!=0A =0ADavid= =0A =0A =0AOn Sunday, June 15, 2025 08:00, "Frantisek Borsik via Cake" said:=0A=0A=0A=0A=0AHello to all,=0A=0AWe're excit= ed to announce the release of the LibreQoS Bufferbloat Test =E2=80=93 an op= en-source bufferbloat testing solution designed specifically for ISPs and n= etwork operators to deploy for their customers.=0A=0ALink=0A[ https://test.= libreqos.com ]( https://test.libreqos.com )=0A=0AWhat Makes This Different= =0A=0AWhile there are several bufferbloat testing tools available, this pla= tform addresses a critical gap: ISP-deployable infrastructure that provides= both traditional testing and realistic household simulation.=0A=0AAs Dave = T=C3=A4ht highlighted in his influential article [ "What's Wrong with Speed= Tests" ]( https://blog.cerowrt.org/post/speedtests/ ), traditional speed t= ests fail to measure what users actually experience. We tried to address Da= ve's points to make a better speed test that focuses on the metric that mat= ters: latency under load in realistic usage scenarios.=0A=0ATwo Complementa= ry Test Modes=0A=0ASingle User Test Mode=0A=0ATraditional sequential load t= esting (baseline =E2=86=92 download =E2=86=92 upload =E2=86=92 bidirectiona= l)=0AMeasures working latency and jitter during each phase=0AFamiliar A+ to= F grading based on latency under load increases=0AComparable to existing t= ools like DSLReports Speed Test and Waveform Bufferbloat Test=0AVirtual Hou= sehold Mode (The Innovation)=0A=0AProcess-isolated simulation of 4 concurre= nt users with authentic traffic patterns:=0A=0AAlex (Gaming): 1.5 Mbps cons= tant, jitter-sensitive for competitive gaming=0ASarah (Video Conference): 2= .5 Mbps bidirectional, Teams simulation with working latency monitoring=0AJ= ake (Netflix HD): 25 Mbps bursts (1s on, 4s off), realistic streaming patte= rns=0AComputer (Background): Up to 200 Mbps continuous download, system upd= ates=0AReal-world relevance: Tests latency under load when multiple family = members are online simultaneously=0A=0AAdvanced grading: Network fairness, = jitter measurement, and per-user working latency analysis=0A=0AWhy (not onl= y) ISPs Need This=0A=0AThe traditional approach of sending customers to thi= rd-party speed test sites has limitations:=0A=0ANo control over test method= ology or server placement=0ALimited correlation with customer support ticke= ts=0AGeneric results that don't reflect real-world usage patterns=0ANo inte= gration with ISP operational systemsThis platform enables (not only) ISPs t= o:=0A=0AHost their own testing infrastructure with full control=0AIntegrate= with support systems via telemetry APIs=0AProvide customers with realistic= household testing scenarios=0ACorrelate test results with network performa= nce and customer complaints=0AOpen Source & Community=0A=0AThe entire platf= orm is open source and available here: [ https://github.com/LibreQoE/buffer= bloat_test ]( https://github.com/LibreQoE/bufferbloat_test )=0A=0AWe've des= igned this to be:=0A=0AEasy to deploy for (not only) ISPs of any size=0ASci= entifically meaningful in its measurement methodology=0ARealistic in its si= mulation of actual household usage=0AIntegrable with existing ISP operation= al workflows=0ACommunity Feedback Requested=0A=0AWe'd love feedback from th= e [ bufferbloat.net ]( http://bufferbloat.net ) community on:=0A=0ATest met= hodology: Are we measuring the right metrics?=0AGrading thresholds: Do our = A+ to F grades align with real-world impact?=0AVirtual household scenarios:= What other realistic usage patterns should we simulate?=0AISP adoption: Wh= at barriers exist for ISP deployment?=0ATechnical Discussion=0A=0AWe'd welc= ome discussion about:=0A=0AMeasurement accuracy for working latency and jit= ter in virtual household mode=0ATraffic pattern authenticity (gaming, video= conferencing, streaming)=0AGrading methodology for latency under load in c= omplex multi-user scenarios=0AIntegration approaches for ISP operational sy= stems=0AThe platform represents our attempt to bridge the gap between acade= mic bufferbloat research and practical ISP operations, building on the foun= dational work of researchers like Dave T=C3=A4ht and the broader bufferbloa= t community. We believe that widespread ISP deployment of proper bufferbloa= t testing infrastructure will ultimately benefit the entire internet ecosys= tem.=0A=0ALooking forward to the community's thoughts and feedback!=0A=0ABe= st regards,=0A=0AThe LibreQoS Team=0A=0A=0A=0A=0A=0A=0A=0A=0A=0A=0A=0A =0AI= n loving memory of Dave T=C3=A4ht: 1965-2025=0A[ https://libreqos.io/2025/0= 4/01/in-loving-memory-of-dave/ ]( https://libreqos.io/2025/04/01/in-loving-= memory-of-dave/ )=0A =0A[ https://www.linkedin.com/in/frantisekborsik ]( ht= tps://www.linkedin.com/in/frantisekborsik )=0ASignal, Telegram, WhatsApp: += 421919416714 =0AiMessage, mobile: +420775230885=0ASkype: casioa5302ca=0A[ f= rantisek.borsik@gmail.com ]( mailto:frantisek.borsik@gmail.com ) ------=_20250616150339000000_79771 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable

I plan to test this ou= t as "being my own ISP" using one of my personal cloud servers that is in o= ne of Vultur's hosting sites as the server. (and from various mobile device= s as I travel around among WiFi sites, if I can).

=0A

The thing I immediately want to know, t= hough, is: has anybody done a serious security review of the system? I'm a = "trust but verify" kind of personality. I don't have the energy or time to = do the detailed review. I trust that the folks who built this are not black= hats - you're my "friends" after all - we care about similar things. But I= 'd like to verify that trust.

=0A

 

=0A

Also, any suggestions people might have with regard to maki= ng it work?

=0A

 

=0A

I'= ll send feedback as I explore, promise! Great work!

=0A

 

=0A

David

=0A

&nbs= p;

=0A

 

=0A

On Sunday, = June 15, 2025 08:00, "Frantisek Borsik via Cake" <cake@lists.bufferbloat= .net> said:

=0A
=0A
=0A
Hello to all,

We're excited to announce the r= elease of the LibreQoS Bufferbloat Test =E2=80=93 an open-= source bufferbloat testing solution designed specifically for ISPs and netw= ork operators to deploy for their customers.

Link
https://test.libreqos.com=

What Makes This Different

While the= re are several bufferbloat testing tools available, this platform addresses= a critical gap: ISP-deployable infrastructure that provides both tradition= al testing and realistic household simulation.

As Dave T=C3=A4ht= highlighted in his influential article "What's Wrong with Speed Tests", traditional speed t= ests fail to measure what users actually experience. We tried to address Da= ve's points to make a better speed test that focuses on the metric that mat= ters: latency under load in realistic usage scenarios.

T= wo Complementary Test Modes

Single User Test Mode
=0A
=0A=0A
=0A
 ------=_20250616150339000000_79771--