- mat-wg - mailman.ripe.net

RIPE Atlas, RIPEstat and RIS Quarterly Planning Q3 2023
by Robert Kisteleki 28 Jun '23

28 Jun '23

Dear colleagues, The quarterly plans for RIPE Atlas, RIPEstat and RIS can now be found at: https://www.ripe.net/support/documentation/quarterly-planning We hope that by providing these plans, we can better inform the community of the ongoing developments on the tools we operate and get feedback on our work. If you have any input on our plans or want to let us know what you would like to see us work on, please let us know. Kind regards, Robert Kisteleki RIPE NCC

1 0

RIPE86 MAT WG agenda, 2023-05-25 (Thursday)
by Stephen Strowes 25 May '23

25 May '23

Hi folks, This week's MAT WG session will take place on Thursday, 25th of May, starting at 09:00 UTC+2 [07:00 UTC] We received many great submissions! Our current agenda is available at: https://ripe86.ripe.net/programme/meeting-plan/mat-wg/ See you there! -Stephen

2 1

Open House: RIPE IPmap and Infrastructure Geolocation
by Chris Amin 15 May '23

15 May '23

Dear colleagues, This is a reminder that the RIPE NCC Open House: RIPE IPmap and Infrastructure Geolocation is taking place this Friday, 19th May at 09:00 UTC (11:00 CEST). This Open House is for anyone interested in RIPE IPmap, the RIPE NCC's API for getting geolocation data for core Internet infrastructure. It will be a sixty-minute remote session where RIPE NCC staff will share context and information about RIPE IPmap, and invite questions and discussion from the community. Details and free registration can be found at: https://www.ripe.net/participate/meetings/open-house/ripe-ncc-open-house-ri… Kind regards, Chris Amin RIPE NCC

1 0

The Resilience of the Internet in Ukraine - One Year On
by Emile Aben 17 Apr '23

17 Apr '23

Dear Colleagues, A year ago I wrote about the resilience of the Internet in Ukraine a few weeks after the war broke out in the country. In this article I revisit the topic and analyse what has changed after more than a year of war, based on data the RIPE NCC collects through RIPE RIS and RIPE Atlas: https://labs.ripe.net/author/emileaben/the-resilience-of-the-internet-in-uk… kind regards, Emile Aben RIPE NCC

1 0

Contribution submission for MAT WG at RIPE86
by Marco Polverini 10 Apr '23

10 Apr '23

Dear Nina, Massimo and Stephen My name is Marco Polverini, an assistant professor at Sapienza University of Rome. I would like to submit the following proposal to be presented at the next MAT WG. This work has been accepted to be presented at the IEEE NOMS 23 ( https://noms2023.ieee-noms.org/) Attached to this email you can find the original paper (in case you need further details about the topic of the presentation I would like to do). Thanks Marco *Title:* A Digital Twin based Framework to Enable "What-If" Analysis in BGP Optimization *Outline of the speech:* - introduction and motivation: in this talk we aim to introduce a framework for the creation of Digital Twins of Autonomous Systems in order to enable the execution of "what if" analysis in the context of inter domain routing optimization - background on digital twins: this section is to introduce the basic terminology and concepts on Digital Twins - presentation of the proposed framework: in this section we provide an overview of the architecture of the proposed framework and detail the responsibilities of each functional block - proof of concept: a simple proof of concept is presented, with the aim of showing the effectiveness of the proposal - conclusion and future work *Abstract:* Nowadays, inter domain routing optimization is performed based on the so called "Tweak and Pray" approach, which consists in performing changes in the configuration of the BGP protocol without knowing in advance the consequences of such a modification. This is due to the lack of cooperation among Network Operators in the configuration of the BGP to optimize the inter domain routing. Inefficiency in the resource usage, network anomalies and outages are common consequences of wrong configuration changes performed by Network Operators in an attempt to improve the performance of their infrastructures. In this paper we propose a novel framework based on the Digital Twin technology to enable the execution of "what-if" analysis in the context of Traffic Engineering performed by tuning BGP parameters. Such a paradigm shift will allow Network Operators to be aware of the effects of BGP configuration changes before their actual execution. A proof of concept related to the balancing of inbound traffic in an Autonomous System network, based on the use of the AS Path Prepending technique, is realized to validate the feasibility of the proposed approach.

2 1

delegate LIR portal access to specific user for specific resources - was: [address-policy-wg] 2023-02 New Policy Proposal (Minimum Size for IPv4 Temporary Assignments)
by Massimo Candela 07 Apr '23

07 Apr '23

Hi Elvis, On 14/03/2023 08:42, Elvis Daniel Velea wrote: > Hi Massimo, > > thank you for your message. Let me see if I have an answer to some of > your questions/comments. See inline. > > ------- > I also mention below a variation of this proposal, which could be it's > own proposal/thread (suggestions welcome): > > There should be an easy way to do "temporary assignments" (which may or > may not be the correct term in this case) to researchers/developers, > starting from address space of a company which is "sponsoring" the > research. > > > The key part of what I would like to have is the possibility to provide > somebody with access to LIR portal services but limited to a specific > subset of my resources. > > In general, a company is not going to support a research/experiment by > providing indiscriminate access to the LIR portal. Creating a new > LIR or > transferring prefixes is not a plausible solution in this context. > > > An existing LIR can do an assignment to a researcher/developer as we > speak. All assignments an LIR makes are ‘temporary’, some may last a day > and some may last 10 years… Can you explain this better? I have been asking to several ncc contacts and there was no suitable solution for this at the moment. Asking researchers/developers to create their own LIRs so that resources can be transferred is not a practical solution, but I'm not sure you are referring to this. Giving the researcher/developer access to the LIR portal/APIs for all the resources is also not a solution. > > If you want the researcher to have access to services like RPKI, they > can ask their LIR. In some cases, maybe temporary transfers could also > be of use. Again, I don't see a way to do this at the moment. > > Note that this proposal aims to update the temporary assignment proposal > that is currently in place. The RIPE NCC makes these temporary > assignments from a pool of IPs they have reserved specifically for this > purpose. I understand. This is why I said that this could be a different proposal (even if related by a similar purpose). I changed the email subject to avoid confusion. > > > Also, I believe this would remove the need for an approval procedure > from the RIPE NCC side: (1) if the address space used is of a company, > there is less need to validate the research project motivations; and > (2) > the company "sponsoring" is also the one responsible for the address > space. > > > I am not sure I understand what you mean by this. The NCC does not need > to approve any assignments made by LIRs. The NCC will need to approve > temporary assignments if the request is sent by an LIR (for an end-user) > based on the current temporary assignment policy. That's exactly what I'm saying. Assignments done by LIRs of their own resources do not require approval, while they would satisfy the research use case (in summary: using prefixes from the ncc stash is not needed in all cases). Thanks for your answers. Ciao, Massimo

1 0

Call for presentations for MAT WG at RIPE86 in May
by Stephen Strowes 06 Apr '23

06 Apr '23

Hi all, RIPE86 is coming up fast: 22 -- 26 May! We are seeking your presentations for the next MAT WG session during that week. Presentations for the WG session are preferably 10-15min long. Please include in your proposal: * a title * a draft of your proposed presentation, even if just an outline * a paragraph describing the topic, and why the topic is of interest for the community. Interesting topics include (but are not limited to): recent measurement research which has an operational impact; recent operational work that may open avenues to research; projects, tools, or datasets that support either operations or research. Please submit your presentation proposal by Friday 28 April by emailing us at mat-wg-chairs at ripe.net. We will make our selections and post the agenda soon after. RIPE meetings are fully hybrid, and we encourage presentations from anywhere. Sessions are facilitated through meetecho. For our planning, we'd appreciate it if you can indicate whether you are likely to be online or on-site. Please ask if you have any questions regarding the setup on either side. Thanks! Cheers, Nina, Massimo, Stephen, as WG co-chairs

1 1

Subject: RIPE Atlas, RIPEstat and RIS Quarterly Planning Q2 2023
by Robert Kisteleki 30 Mar '23

30 Mar '23

Dear colleagues, The quarterly plans for RIPE Atlas, RIPEstat and RIS can now be found at: https://www.ripe.net/support/documentation/quarterly-planning The goal of our quarterly planning is to communicate the work we’re doing for the coming months, get your input and start a discussion on other developments we should work on. If you have any input on our plans or want to let us know what you would like to see us work on, please let us know. Kind regards, Robert Kisteleki RIPE NCC

1 0

Re: [mat-wg] [address-policy-wg] 2023-02 New Policy Proposal (Minimum Size for IPv4 Temporary Assignments)
by Massimo Candela 07 Mar '23

07 Mar '23

Hi all, I added in cc the MAT mailing list, which is frequented by researchers. I believe that is the right audience for this proposal. @MAT: Support for this proposal should be expressed before this Friday (March 10). See proposal below. Reply by keeping both list. @Addres-policy: I support this proposal. ------- I also mention below a variation of this proposal, which could be it's own proposal/thread (suggestions welcome): There should be an easy way to do "temporary assignments" (which may or may not be the correct term in this case) to researchers/developers, starting from address space of a company which is "sponsoring" the research. The key part of what I would like to have is the possibility to provide somebody with access to LIR portal services but limited to a specific subset of my resources. In general, a company is not going to support a research/experiment by providing indiscriminate access to the LIR portal. Creating a new LIR or transferring prefixes is not a plausible solution in this context. Also, I believe this would remove the need for an approval procedure from the RIPE NCC side: (1) if the address space used is of a company, there is less need to validate the research project motivations; and (2) the company "sponsoring" is also the one responsible for the address space. Ciao, Massimo On 27/02/2023 16:44, Leo Vegoda wrote: > Hi, > > We've not had any feedback on this proposal yet. > > As a reminder, this proposal would set "the minimum assignment size to > a /24 while still allowing for a smaller assignment if requested by > the End User. This policy proposal also allows routing requirements to > justify the request for more than a /24 for research purposes." > > Support for the proposal, or arguments against it are welcome. > > Many thanks, > > Leo Vegoda > Address Policy WG co-chair > > On Thu, 9 Feb 2023 at 00:26, Angela Dall'Ara <adallara(a)ripe.net> wrote: >> >> Dear colleagues, >> >> A new RIPE Policy Proposal, 2023-02, "Minimum Size for IPv4 Temporary >> Assignments" >> is now available for discussion. >> >> This policy proposal recommends setting the minimum assignment size for >> temporary assignments to a /24 while still allowing for a smaller >> assignment if requested by the End User. This policy proposal also >> allows routing requirements to justify the request for more than a /24 >> for research purposes. >> >> You can find the full proposal at: >> https://www.ripe.net/participate/policies/proposals/2023-02 >> >> As per the RIPE Policy Development Process (PDP), the purpose of this >> four-week Discussion Phase is to discuss the proposal and provide >> feedback to the proposers. >> >> At the end of the Discussion Phase, the proposers, with the agreement of >> the WG Chairs, will decide how to proceed with the proposal. >> >> The PDP document can be found at: >> https://www.ripe.net/publications/docs/ripe-781 >> >> We encourage you to review this proposal and send your comments to >> address-policy-wg(a)ripe.net before 10 March 2023. >> >> >> Kind regards, >> >> Angela Dall'Ara >> Policy Officer >> RIPE NCC >> >> >> -- >> >> To unsubscribe from this mailing list, get a password reminder, or change your subscription options, please visit: https://lists.ripe.net/mailman/listinfo/address-policy-wg >

1 0

Improved timeliness of RIS dump files
by Emile Aben 07 Mar '23

07 Mar '23

Dear colleagues, We are planning to replace the pipeline that generates MRT dump files publicly available from the RIPE NCC. The new pipeline will have significantly less delay, and also some changes on the structure of the MRT file format [*]. Files in the newer format are available for testing here: https://www.ris.ripe.net/dumps-per-rrc-rest/prototype/ We have received feedback from MRT parser implementers that these files work fine for their parsers, but in case you haven't tested your MRT parser implementation, we ask you to test the new files before we make the switch. Our plan is to make this switch on 3 April 2023. After we make the switch the bview files available at our main location for downloading MRT files will be generated using the new pipeline and we will be able to produce these files in a more timely manner. Kind regards, Emile Aben RIPE NCC [*] These changes are documented here: https://ris.ripe.net/docs/40_Prototypes/15_per_rrc_dumps.html, but beware at this point we are only changing the files, not the frequency of dump files, nor putting the API in production. We did not receive feedback about these, but please let us know if you find them useful.

1 0

Re: [mat-wg] [atlas] Proposal: Remove support for non-public measurements [ONLY-PUBLIC]
by Stephen Strowes 11 Jan '23

11 Jan '23

Hi, On 12/31/22 18:37, Randy Bush wrote: > hi robert, > > and thanks > >> I seem to remember that a typical non-public measurement is a one-off, >> and as such the data volume collected and stored is likely to be >> relatiely low. > interesting There is an old (late 2016) labs post worth referring to for this discussion [1]. In 2020 there was a survey of the users of private measurements [2], which was intended to get feedback on their actual use-cases. I seem to recall a very small set of respondents from an already small set of surveyees. The results are probably lost in the ether. I feel there's an aspect of the broader question that is rarely touched upon in these discussions: would user behaviour actually change if non-public measurements went away? I'm never sure how many use-cases are truly secret. Cheers, S. [1] https://labs.ripe.net/author/kistel/non-public-measurements-in-ripe-atlas/ [2] https://www.ripe.net/participate/forms/apply/ripe-atlas-private-measurement… > >> I don't have the numbers readily available to me at this moment, but >> we can certainly dig up recent statistics about this. > i did not intend to create a lot of work for you. but, as a measurement > person, i am a bit fond of basing decisions on facts. > > randy >

1 0

Fwd: [atlas] Proposal: Remove support for non-public measurements [ONLY-PUBLIC]
by Stephen Strowes 22 Dec '22

22 Dec '22

Hi all, I'm forwarding these five threads into the MAT WG because they're all probably of interest to folks here. Please share your feedback with the NCC! Cheers, S. -------- Forwarded Message -------- Subject: [atlas] Proposal: Remove support for non-public measurements [ONLY-PUBLIC] Date: Wed, 14 Dec 2022 15:04:40 +0100 From: Robert Kisteleki <robert(a)ripe.net> Organization: RIPE NCC To: ripe-atlas(a)ripe.net <ripe-atlas(a)ripe.net> Dear RIPE Atlas users, We recently published a RIPE Labs article containing a few proposals: https://labs.ripe.net/author/kistel/five-proposals-for-a-better-ripe-atlas/. We'd like to encourage you to express your comments about this proposal (if you'd like to share them) here. Regards, Robert Kisteleki For the RIPE Atlas team -- ripe-atlas mailing list ripe-atlas(a)ripe.net https://lists.ripe.net/mailman/listinfo/ripe-atlas

2 1

Re: [mat-wg] [atlas] Proposal: Measure well-known CDNs,[CDN-HTTP]
by Massimo Candela 17 Dec '22

17 Dec '22

On 17/12/2022 00:19, Barry Raveendran Greene wrote: > > >> On Dec 16, 2022, at 23:29, Stephane Bortzmeyer <bortzmeyer(a)nic.fr> wrote: >> >> There is a larger problem here, a more strategic one: such a feature >> would contribute to the centralisation of the Internet, which is >> already too important. Tagging some targets are "important" and >> "worthy of measurements" would mean that we consider some HTTP servers >> to be more useful than others. That would be a bad message from RIPE. > > We’ve come full circle - we started with centralized PTTs - moved to a decentralized ASN/Paul Baran model - now Re-centralized based on marketing domination. > > +1 with Stephane’s observation. The selection of who to measure is a statement. +1 Also, while the data would be useful, I don't think the role of the ripe ncc is to grade commercial services. Let other companies or individual researchers do that.

1 0

RIPE Atlas, RIPEstat and RIS Quarterly Planning Q1 2023
by Robert Kisteleki 16 Dec '22

16 Dec '22

Dear colleagues, We’ve now published our quarterly planning for RIPE Atlas, RIPEstat and RIS. These are available at: https://www.ripe.net/support/documentation/quarterly-planning The goal of our quarterly planning is to communicate the work we’re doing for the coming months, get your input and start a discussion on other developments we should work on. If you have any input on our plans or want to let us know what you would like to see us work on, please let us know. Kind regards, Robert Kisteleki Research and Development Manager RIPE NCC

1 0

Fwd: [atlas] Proposal: Add support for STARTTLS measurements [STARTTLS]
by Stephen Strowes 14 Dec '22

14 Dec '22

Hi all, I'm forwarding these five threads into the MAT WG because they're all probably of interest to folks here. Please share your feedback with the NCC! Cheers, S. -------- Forwarded Message -------- Subject: [atlas] Proposal: Add support for STARTTLS measurements [STARTTLS] Date: Wed, 14 Dec 2022 15:04:06 +0100 From: Robert Kisteleki <robert(a)ripe.net> Organization: RIPE NCC To: ripe-atlas(a)ripe.net <ripe-atlas(a)ripe.net> Dear RIPE Atlas users, We recently published a RIPE Labs article containing a few proposals: https://labs.ripe.net/author/kistel/five-proposals-for-a-better-ripe-atlas/. We'd like to encourage you to express your comments about this proposal (if you'd like to share them) here. Regards, Robert Kisteleki For the RIPE Atlas team -- ripe-atlas mailing list ripe-atlas(a)ripe.net https://lists.ripe.net/mailman/listinfo/ripe-atlas

1 0

Fwd: [atlas] Proposal: Generic HTTP measurements [GENERIC-HTTP]
by Stephen Strowes 14 Dec '22

14 Dec '22

Hi all, I'm forwarding these five threads into the MAT WG because they're all probably of interest to folks here. Please share your feedback with the NCC! Cheers, S. -------- Forwarded Message -------- Subject: [atlas] Proposal: Generic HTTP measurements [GENERIC-HTTP] Date: Wed, 14 Dec 2022 15:03:32 +0100 From: Robert Kisteleki <robert(a)ripe.net> Organization: RIPE NCC To: ripe-atlas(a)ripe.net <ripe-atlas(a)ripe.net> Dear RIPE Atlas users, We recently published a RIPE Labs article containing a few proposals: https://labs.ripe.net/author/kistel/five-proposals-for-a-better-ripe-atlas/. We'd like to encourage you to express your comments about this proposal (if you'd like to share them) here. Regards, Robert Kisteleki For the RIPE Atlas team -- ripe-atlas mailing list ripe-atlas(a)ripe.net https://lists.ripe.net/mailman/listinfo/ripe-atlas

1 0

Fwd: [atlas] Proposal: Measure well-known CDNs, [CDN-HTTP]
by Stephen Strowes 14 Dec '22

14 Dec '22

Hi all, I'm forwarding these five threads into the MAT WG because they're all probably of interest to folks here. Please share your feedback with the NCC! Cheers, S. -------- Forwarded Message -------- Subject: [atlas] Proposal: Measure well-known CDNs,[CDN-HTTP] Date: Wed, 14 Dec 2022 15:02:46 +0100 From: Robert Kisteleki <robert(a)ripe.net> Organization: RIPE NCC To: ripe-atlas(a)ripe.net <ripe-atlas(a)ripe.net> Dear RIPE Atlas users, We recently published a RIPE Labs article containing a few proposals: https://labs.ripe.net/author/kistel/five-proposals-for-a-better-ripe-atlas/. We'd like to encourage you to express your comments about this proposal (if you'd like to share them) here. Regards, Robert Kisteleki For the RIPE Atlas team -- ripe-atlas mailing list ripe-atlas(a)ripe.net https://lists.ripe.net/mailman/listinfo/ripe-atlas

1 0

Fwd: [atlas] Proposal: remove per-hop “late” packets from traceroute [LATE-PACKETS]
by Stephen Strowes 14 Dec '22

14 Dec '22

Hi all, I'm forwarding these five threads into the MAT WG because they're all probably of interest to folks here. Please share your feedback with the NCC! Cheers, S. -------- Forwarded Message -------- Subject: [atlas] Proposal: remove per-hop “late” packets from traceroute [LATE-PACKETS] Date: Wed, 14 Dec 2022 15:02:13 +0100 From: Robert Kisteleki <robert(a)ripe.net> Organization: RIPE NCC To: ripe-atlas(a)ripe.net <ripe-atlas(a)ripe.net> Dear RIPE Atlas users, We recently published a RIPE Labs article containing a few proposals: https://labs.ripe.net/author/kistel/five-proposals-for-a-better-ripe-atlas/. We'd like to encourage you to express your comments about this proposal (if you'd like to share them) here. Regards, Robert Kisteleki For the RIPE Atlas team -- ripe-atlas mailing list ripe-atlas(a)ripe.net https://lists.ripe.net/mailman/listinfo/ripe-atlas

1 0

MAT Working Group Minutes - RIPE 85
by Massimo Candela 29 Nov '22

29 Nov '22

Dear all, The MAT Working Group Minutes for RIPE 85 are available at https://www.ripe.net/participate/ripe/wg/active-wg/mat/minutes/mat-working-… Ciao, Massimo (on behalf of the MAT Chairs)

1 0

Format change for index pages of RIS MRT data files and RIPE Atlas daily dataset dumps
by Oleg Muravskiy 24 Nov '22

24 Nov '22

Dear RIS and RIPE Atlas users, On Wednesday, 30 November 2022 we are planning to migrate the web server at <https://data.ris.ripe.net/> as well as <https://data-store.ripe.net> from Apache httpd to nginx. This will affect how the file listings in HTML index pages are generated, including the following: * RIS MRT data files (updates and dumps) * RIPE Atlas daily dataset dumps If you rely on the format of these pages, you need to update your software. Before we do the upgrade, you could check the new format of the index pages at: https://test-data.ris.ripe.net/rrc00/ We would like to ask you to avoid relying on the specific format of the index.html files. If you are fetching BGP dump files, please use the REST API that is described at: https://ris.ripe.net/docs/40_Prototypes/15_per_rrc_dumps.html You might also consider using per-peer dump files instead: https://ris.ripe.net/docs/40_Prototypes/10_per_peer_dumps.html We want to encourage you to provide your feedback on the API, so that we could improve it and build a similar API for the update files. If you are fetching update files, please use direct URLs to files as described at: https://ris.ripe.net/docs/20_raw_data_mrt.html#name-and-location Kind regards, Oleg Muravskiy, Robert Kisteleki RIPE NCC

1 0

streaming link MAT session ripe85
by Massimo Candela 27 Oct '22

27 Oct '22

The MAT session started at RIPE85. Streaming here: https://ripe85.ripe.net/live/side/

1 0

MAT agenda for RIPE85
by Massimo Candela 26 Oct '22

26 Oct '22

Dear all, I'm pleased to announce that we have an agenda for RIPE85! https://ripe85.ripe.net/programme/meeting-plan/mat-wg/ Ciao, Massimo (on behalf of the MAT chairs)

2 1

Re: [mat-wg] [ippm] [Rpm] [M-Lab-Discuss] misery metrics & consequences
by J Ignacio Alvarez-Hamelin 26 Oct '22

26 Oct '22

Dear all, After some time in silence on the IPPM list, I like to make some comments here. As we presented in the draft-ietf-ippm-route-00 (now the RFC9198), the main problem is the traffic follows heavy-tailed distributions when it is seen from the end-to-end points: the origin of most of the issues in that video. Therefore, treating it as parametric distribution is not possible, unless you are dealing with a complex distribution like the Stable distribution: • B. Mandelbrot, “New methods in statistical economics,” Journal of political economy, vol. 71, no. 5, pp. 421–440, 1963. • ——, “The variation of certain speculative prices,” The journal of business, vol. 36, no. 4, pp. 394–419, 1963. (and so it will be extremely complex a high computing demands.) This is why we propose to use quartiles to characterize delays in the RFC9198. Then, I am doing some research to understand how the delay can change with network load, using the quartiles. I attach some measurements done during the pandemic, showing the congestion as a function of the time. Best, Ignacio _______________________________________________________________ Dr. Ing. José Ignacio Alvarez-Hamelin CONICET and Facultad de Ingeniería, Universidad de Buenos Aires Av. Paseo Colón 850 - C1063ACV - Buenos Aires - Argentina +54 (11) 5285 0716 / 5285 0705 e-mail: ihameli(a)cnet.fi.uba.ar web: http://cnet.fi.uba.ar/ignacio.alvarez-hamelin/ _______________________________________________________________ > On 23 Oct 2022, at 08:57, Sebastian Moeller <moeller0(a)gmx.de> wrote: > > Hi Glenn, > > >> On Oct 23, 2022, at 02:17, Glenn Fishbine via Rpm <rpm(a)lists.bufferbloat.net> wrote: >> >> As a classic died in the wool empiricist, granted that you can identify "misery" factors, given a population of 1,000 users, how do you propose deriving a misery index for that population? >> >> We can measure download, upload, ping, jitter pretty much without user intervention. For the measurements you hypothesize, how you you automatically extract those indecies without subjective user contamination. >> >> I.e. my download speed sucks. Measure the download speed. >> >> My isp doesn't fix my problem. Measure what? How? >> >> Human survey technology is 70+ years old and it still has problems figuring out how to correlate opinion with fact. >> >> Without an objective measurement scheme that doesn't require human interaction, the misery index is a cool hypothesis with no way to link to actual data. What objective measurements can be made? Answer that and the index becomes useful. Otherwise it's just consumer whining. >> >> Not trying to be combative here, in fact I like the concept you support, but I'm hard pressed to see how the concept can lead to data, and the data lead to policy proposals. > > [SM] So it seems that outside of seemingly simple to test throughput numbers*, the next most important quality number (or the most important depending on subjective ranking) is how does latency change under "load". Absolute latency is also important albeit static high latency can be worked around within limits so the change under load seems more relevant. > All of flent's RRUL test, apple's networkQuality/RPM, and iperf2's bounceback test offer methods to asses latency change under load**, as do waveforms bufferbloat tests and even to a degree Ookla's speedtest.net. IMHO something like latency increase under load or apple's responsiveness measure RPM (basically the inverse of the latency under load calculated on a per minute basis, so it scales in the typical higher numbers are better way, unlike raw latency under load numbers where smaller is better). > IMHO what networkQuality is missing ATM is to measure and report the unloaded RPM as well as the loaded the first gives a measure over the static latency the second over how well things keep working if capacity gets tight. They report the base RTT which can be converted to RPM. As an example: > > macbook:~ user$ networkQuality -v > ==== SUMMARY ==== > Upload capacity: 24.341 Mbps > Download capacity: 91.951 Mbps > Upload flows: 20 > Download flows: 16 > Responsiveness: High (2123 RPM) > Base RTT: 16 > Start: 10/23/22, 13:44:39 > End: 10/23/22, 13:44:53 > OS Version: Version 12.6 (Build 21G115) > > Here RPM 2133 corresponds to 60000/2123 = 28.26 ms latency under load, while the Base RTT of 16ms corresponds to 60000/16 = 3750 RPM, son on this link load reduces the responsiveness by 3750-2123 = 1627 RPM a reduction by 100-100*2123/3750 = 43.4%, and that is with competent AQM and scheduling on the router. > > Without competent AQM/shaping I get: > ==== SUMMARY ==== > Upload capacity: 15.101 Mbps > Download capacity: 97.664 Mbps > Upload flows: 20 > Download flows: 12 > Responsiveness: Medium (427 RPM) > Base RTT: 16 > Start: 10/23/22, 13:51:50 > End: 10/23/22, 13:52:06 > OS Version: Version 12.6 (Build 21G115) > latency under load: 60000/427 = 140.52 ms > base RPM: 60000/16 = 3750 RPM > reduction RPM: 100-100*427/3750 = 88.6% > > > I understand apple's desire to have a single reported number with a single qualifier medium/high/... because in the end a link is only reliably usable if responsiveness under load stays acceptable, but with two numbers it is easier to see what one's ISP could do to help. (I guess some ISPs might already be unhappy with the single number, so this needs some diplomacy/tact) > > Regards > Sebastian > > > > *) Seemingly as quite some ISPs operate their own speedtest servers in their network and ignore customers not reaching the contracted rates into speedtest-servers located in different ASs. As the product is called internet access I a inclined to expect that my ISP maintains sufficient peering/transit capacity to reach the next tier of AS at my contracted rate (the EU legislative seems to agree, see EU directive 2015/2120). > > **) Most do by creating load themselves and measuring throughput at the same time, bounceback IIUC will focus on the latency measurement and leave the load generation optional (so offers a mode to measure responsiveness of a live network with minimal measurement traffic). @Bob, please correct me if this is wrong. > > >> >> >> On Fri, Oct 21, 2022, 5:20 PM Dave Taht <dave.taht(a)gmail.com> wrote: >> One of the best talks I've ever seen on how to measure customer >> satisfaction properly just went up after the P99 Conference. >> >> It's called Misery Metrics. >> >> After going through a deep dive as to why and how we think and act on >> percentiles, bins, and other statistical methods as to how we use the >> web and internet are *so wrong* (well worth watching and thinking >> about if you are relying on or creating network metrics today), it >> then points to the real metrics that matter to users and the ultimate >> success of an internet business: Timeouts, retries, misses, failed >> queries, angry phone calls, abandoned shopping carts and loss of >> engagement. >> >> https://www.p99conf.io/session/misery-metrics-consequences/ >> >> The ending advice was - don't aim to make a specific percentile >> acceptable, aim for an acceptable % of misery. >> >> I enjoyed the p99 conference more than any conference I've attended in years. >> >> -- >> This song goes out to all the folk that thought Stadia would work: >> https://www.linkedin.com/posts/dtaht_the-mushroom-song-activity-69813666656… >> Dave Täht CEO, TekLibre, LLC >> >> -- >> You received this message because you are subscribed to the Google Groups "discuss" group. >> To unsubscribe from this group and stop receiving emails from it, send an email to discuss+unsubscribe(a)measurementlab.net. >> To view this discussion on the web visit https://groups.google.com/a/measurementlab.net/d/msgid/discuss/CAA93jw4w27a…. >> _______________________________________________ >> Rpm mailing list >> Rpm(a)lists.bufferbloat.net >> https://lists.bufferbloat.net/listinfo/rpm > > _______________________________________________ > ippm mailing list > ippm(a)ietf.org > https://www.ietf.org/mailman/listinfo/ippm

2 1

Re: [mat-wg] [Rpm] [ippm] [M-Lab-Discuss] misery metrics & consequences
by J Ignacio Alvarez-Hamelin 25 Oct '22

25 Oct '22

Dear Bob, I answer inline. Thank you for your comments. P.D. It seems that I have no rights to send emails to the Measurement Analysis and Tools Working Group: how I can join it? _______________________________________________________________ Dr. Ing. José Ignacio Alvarez-Hamelin CONICET and Facultad de Ingeniería, Universidad de Buenos Aires Av. Paseo Colón 850 - C1063ACV - Buenos Aires - Argentina +54 (11) 5285 0716 / 5285 0705 e-mail: ihameli(a)cnet.fi.uba.ar web: http://cnet.fi.uba.ar/ignacio.alvarez-hamelin/ _______________________________________________________________ > On 25 Oct 2022, at 14:02, rjmcmahon <rjmcmahon(a)rjmcmahon.com> wrote: > > I don't understand the information in the link. It looks like lines on a map to some form of dials which are too small to read. I suppose that depends on the zoom of your browser, try to put 100% (command+ to zoom-in in chrome). You will see a 24 hour clock where some hours are green (no congestion), yellow (some congestion) or red (congestion). But all of that is just an example. > > One can sample and create a Gaussian per the central limit theorem (CLT) if the underlying process probability density functions converge, i.e. can be integrated to 1. With that said, normalizing does lose information and doesn't say much about tails, outliers, etc. The Stable distribution is a generalization over all the distributions. Consider that Gaussian has finite moments, and heavy-tailed distribution not, this is where Stable one enters as a generalization. > > We should be careful in assuming only the tails matter and that all traffic follows heavy-tailed distributions. With bufferbloat it's the minimum of the latency PDF that shifts. Codel watches a minimum. "Jacobson suggested that average queue length actually contains no information at all about packet demand or network load.[3][5] He suggested that a better metric might be the minimum queue length during a sliding time window." Well, it depends on which point of the network you really are, but there a lot where this is true. In the other hand, no heavy tailed distributions do not cause problems if you use statistical tools that suport anything. For example, the median (Q2) converges to the mean (fist statistical moment) for Poisson distributions. > > We need statistical tools that also allow for the analysis of non-parametric distributions too. Hotelling T2 assumes the multivariate distributions are Gaussian. Kolmogorov-Smirnov tests can be used for non-parametric distributions. We find both are needed for SPC used by our automation systems. Yes, I did a lot of this kind of test, and I found very interesting things, even in the very important links (important ISP with important CDN), verifying Stable distributions with certain parameters (even some ones with heavy-tailed distributions). The problem is how to do this a low computable cost, and this why you can use quartiles. > > Sample subgroups of one really don't give sufficient information about any type of distribution, parametric or non-parametric. Here I cannot understand: which are the subgroups? > > Bob >> Dear all, >> After some time in silence on the IPPM list, I like to make some >> comments here. As we presented in the draft-ietf-ippm-route-00 (now >> the RFC9198), the main problem is the traffic follows heavy-tailed >> distributions when it is seen from the end-to-end points: the origin >> of most of the issues in that video. Therefore, treating it as >> parametric distribution is not possible, unless you are dealing with a >> complex distribution like the Stable distribution: >> • B. Mandelbrot, “New methods in statistical economics,” Journal of >> political economy, vol. 71, no. 5, pp. 421–440, 1963. >> • ——, “The variation of certain speculative prices,” The journal of >> business, vol. 36, no. 4, pp. 394–419, 1963. >> (and so it will be extremely complex a high computing demands.) >> This is why we propose to use quartiles to characterize delays in the >> RFC9198. Then, I am doing some research to understand how the delay >> can change with network load, using the quartiles. >> You can see some measurements done during the pandemic, showing the >> congestion as a function of the time (24 hours maximum): >> https://cnet.fi.uba.ar/ignacio.alvarez-hamelin/RIPE-Atlas-measurement-24681… >> [you can zoom in and out, pan it, clicking on the Xs you can close >> dialogs, to reopen them click on the link] >> Best, >> Ignacio >> ___________________________________ >> _______________________________________________________________ >> Dr. Ing. José Ignacio Alvarez-Hamelin >> CONICET and Facultad de Ingeniería, Universidad de Buenos Aires >> Av. Paseo Colón 850 - C1063ACV - Buenos Aires - Argentina >> +54 (11) 5285 0716 / 5285 0705 >> e-mail: ihameli(a)cnet.fi.uba.ar >> web: http://cnet.fi.uba.ar/ignacio.alvarez-hamelin/ >> _______________________________________________________________ >>> On 23 Oct 2022, at 08:57, Sebastian Moeller <moeller0(a)gmx.de> wrote: >>> Hi Glenn, >>>> On Oct 23, 2022, at 02:17, Glenn Fishbine via Rpm <rpm(a)lists.bufferbloat.net> wrote: >>>> As a classic died in the wool empiricist, granted that you can identify "misery" factors, given a population of 1,000 users, how do you propose deriving a misery index for that population? >>>> We can measure download, upload, ping, jitter pretty much without user intervention. For the measurements you hypothesize, how you you automatically extract those indecies without subjective user contamination. >>>> I.e. my download speed sucks. Measure the download speed. >>>> My isp doesn't fix my problem. Measure what? How? >>>> Human survey technology is 70+ years old and it still has problems figuring out how to correlate opinion with fact. >>>> Without an objective measurement scheme that doesn't require human interaction, the misery index is a cool hypothesis with no way to link to actual data. What objective measurements can be made? Answer that and the index becomes useful. Otherwise it's just consumer whining. >>>> Not trying to be combative here, in fact I like the concept you support, but I'm hard pressed to see how the concept can lead to data, and the data lead to policy proposals. >>> [SM] So it seems that outside of seemingly simple to test throughput numbers*, the next most important quality number (or the most important depending on subjective ranking) is how does latency change under "load". Absolute latency is also important albeit static high latency can be worked around within limits so the change under load seems more relevant. >>> All of flent's RRUL test, apple's networkQuality/RPM, and iperf2's bounceback test offer methods to asses latency change under load**, as do waveforms bufferbloat tests and even to a degree Ookla's speedtest.net. IMHO something like latency increase under load or apple's responsiveness measure RPM (basically the inverse of the latency under load calculated on a per minute basis, so it scales in the typical higher numbers are better way, unlike raw latency under load numbers where smaller is better). >>> IMHO what networkQuality is missing ATM is to measure and report the unloaded RPM as well as the loaded the first gives a measure over the static latency the second over how well things keep working if capacity gets tight. They report the base RTT which can be converted to RPM. As an example: >>> macbook:~ user$ networkQuality -v >>> ==== SUMMARY ==== >>> Upload capacity: 24.341 Mbps >>> Download capacity: 91.951 Mbps >>> Upload flows: 20 >>> Download flows: 16 >>> Responsiveness: High (2123 RPM) >>> Base RTT: 16 >>> Start: 10/23/22, 13:44:39 >>> End: 10/23/22, 13:44:53 >>> OS Version: Version 12.6 (Build 21G115) >>> Here RPM 2133 corresponds to 60000/2123 = 28.26 ms latency under load, while the Base RTT of 16ms corresponds to 60000/16 = 3750 RPM, son on this link load reduces the responsiveness by 3750-2123 = 1627 RPM a reduction by 100-100*2123/3750 = 43.4%, and that is with competent AQM and scheduling on the router. >>> Without competent AQM/shaping I get: >>> ==== SUMMARY ==== >>> Upload capacity: 15.101 Mbps >>> Download capacity: 97.664 Mbps >>> Upload flows: 20 >>> Download flows: 12 >>> Responsiveness: Medium (427 RPM) >>> Base RTT: 16 >>> Start: 10/23/22, 13:51:50 >>> End: 10/23/22, 13:52:06 >>> OS Version: Version 12.6 (Build 21G115) >>> latency under load: 60000/427 = 140.52 ms >>> base RPM: 60000/16 = 3750 RPM >>> reduction RPM: 100-100*427/3750 = 88.6% >>> I understand apple's desire to have a single reported number with a single qualifier medium/high/... because in the end a link is only reliably usable if responsiveness under load stays acceptable, but with two numbers it is easier to see what one's ISP could do to help. (I guess some ISPs might already be unhappy with the single number, so this needs some diplomacy/tact) >>> Regards >>> Sebastian >>> *) Seemingly as quite some ISPs operate their own speedtest servers in their network and ignore customers not reaching the contracted rates into speedtest-servers located in different ASs. As the product is called internet access I a inclined to expect that my ISP maintains sufficient peering/transit capacity to reach the next tier of AS at my contracted rate (the EU legislative seems to agree, see EU directive 2015/2120). >>> **) Most do by creating load themselves and measuring throughput at the same time, bounceback IIUC will focus on the latency measurement and leave the load generation optional (so offers a mode to measure responsiveness of a live network with minimal measurement traffic). @Bob, please correct me if this is wrong. >>>> On Fri, Oct 21, 2022, 5:20 PM Dave Taht <dave.taht(a)gmail.com> wrote: >>>> One of the best talks I've ever seen on how to measure customer >>>> satisfaction properly just went up after the P99 Conference. >>>> It's called Misery Metrics. >>>> After going through a deep dive as to why and how we think and act on >>>> percentiles, bins, and other statistical methods as to how we use the >>>> web and internet are *so wrong* (well worth watching and thinking >>>> about if you are relying on or creating network metrics today), it >>>> then points to the real metrics that matter to users and the ultimate >>>> success of an internet business: Timeouts, retries, misses, failed >>>> queries, angry phone calls, abandoned shopping carts and loss of >>>> engagement. >>>> https://www.p99conf.io/session/misery-metrics-consequences/ >>>> The ending advice was - don't aim to make a specific percentile >>>> acceptable, aim for an acceptable % of misery. >>>> I enjoyed the p99 conference more than any conference I've attended in years. >>>> -- >>>> This song goes out to all the folk that thought Stadia would work: >>>> https://www.linkedin.com/posts/dtaht_the-mushroom-song-activity-69813666656… >>>> Dave Täht CEO, TekLibre, LLC >>>> -- >>>> You received this message because you are subscribed to the Google Groups "discuss" group. >>>> To unsubscribe from this group and stop receiving emails from it, send an email to discuss+unsubscribe(a)measurementlab.net. >>>> To view this discussion on the web visit https://groups.google.com/a/measurementlab.net/d/msgid/discuss/CAA93jw4w27a…. >>>> _______________________________________________ >>>> Rpm mailing list >>>> Rpm(a)lists.bufferbloat.net >>>> https://lists.bufferbloat.net/listinfo/rpm >>> _______________________________________________ >>> ippm mailing list >>> ippm(a)ietf.org >>> https://www.ietf.org/mailman/listinfo/ippm >> _______________________________________________ >> Rpm mailing list >> Rpm(a)lists.bufferbloat.net >> https://lists.bufferbloat.net/listinfo/rpm

1 0

Re: [mat-wg] [ippm] [Rpm] [M-Lab-Discuss] misery metrics & consequences
by J Ignacio Alvarez-Hamelin 25 Oct '22

25 Oct '22

Dear all, After some time in silence on the IPPM list, I like to make some comments here. As we presented in the draft-ietf-ippm-route-00 (now the RFC9198), the main problem is the traffic follows heavy-tailed distributions when it is seen from the end-to-end points: the origin of most of the issues in that video. Therefore, treating it as parametric distribution is not possible, unless you are dealing with a complex distribution like the Stable distribution: • B. Mandelbrot, “New methods in statistical economics,” Journal of political economy, vol. 71, no. 5, pp. 421–440, 1963. • ——, “The variation of certain speculative prices,” The journal of business, vol. 36, no. 4, pp. 394–419, 1963. (and so it will be extremely complex a high computing demands.) This is why we propose to use quartiles to characterize delays in the RFC9198. Then, I am doing some research to understand how the delay can change with network load, using the quartiles. You can see some measurements done during the pandemic, showing the congestion as a function of the time (24 hours maximum): https://cnet.fi.uba.ar/ignacio.alvarez-hamelin/RIPE-Atlas-measurement-24681… [you can zoom in and out, pan it, clicking on the Xs you can close dialogs, to reopen them click on the link] Best, Ignacio ___________________________________ _______________________________________________________________ Dr. Ing. José Ignacio Alvarez-Hamelin CONICET and Facultad de Ingeniería, Universidad de Buenos Aires Av. Paseo Colón 850 - C1063ACV - Buenos Aires - Argentina +54 (11) 5285 0716 / 5285 0705 e-mail: ihameli(a)cnet.fi.uba.ar web: http://cnet.fi.uba.ar/ignacio.alvarez-hamelin/ _______________________________________________________________ > On 23 Oct 2022, at 08:57, Sebastian Moeller <moeller0(a)gmx.de> wrote: > > Hi Glenn, > > >> On Oct 23, 2022, at 02:17, Glenn Fishbine via Rpm <rpm(a)lists.bufferbloat.net> wrote: >> >> As a classic died in the wool empiricist, granted that you can identify "misery" factors, given a population of 1,000 users, how do you propose deriving a misery index for that population? >> >> We can measure download, upload, ping, jitter pretty much without user intervention. For the measurements you hypothesize, how you you automatically extract those indecies without subjective user contamination. >> >> I.e. my download speed sucks. Measure the download speed. >> >> My isp doesn't fix my problem. Measure what? How? >> >> Human survey technology is 70+ years old and it still has problems figuring out how to correlate opinion with fact. >> >> Without an objective measurement scheme that doesn't require human interaction, the misery index is a cool hypothesis with no way to link to actual data. What objective measurements can be made? Answer that and the index becomes useful. Otherwise it's just consumer whining. >> >> Not trying to be combative here, in fact I like the concept you support, but I'm hard pressed to see how the concept can lead to data, and the data lead to policy proposals. > > [SM] So it seems that outside of seemingly simple to test throughput numbers*, the next most important quality number (or the most important depending on subjective ranking) is how does latency change under "load". Absolute latency is also important albeit static high latency can be worked around within limits so the change under load seems more relevant. > All of flent's RRUL test, apple's networkQuality/RPM, and iperf2's bounceback test offer methods to asses latency change under load**, as do waveforms bufferbloat tests and even to a degree Ookla's speedtest.net. IMHO something like latency increase under load or apple's responsiveness measure RPM (basically the inverse of the latency under load calculated on a per minute basis, so it scales in the typical higher numbers are better way, unlike raw latency under load numbers where smaller is better). > IMHO what networkQuality is missing ATM is to measure and report the unloaded RPM as well as the loaded the first gives a measure over the static latency the second over how well things keep working if capacity gets tight. They report the base RTT which can be converted to RPM. As an example: > > macbook:~ user$ networkQuality -v > ==== SUMMARY ==== > Upload capacity: 24.341 Mbps > Download capacity: 91.951 Mbps > Upload flows: 20 > Download flows: 16 > Responsiveness: High (2123 RPM) > Base RTT: 16 > Start: 10/23/22, 13:44:39 > End: 10/23/22, 13:44:53 > OS Version: Version 12.6 (Build 21G115) > > Here RPM 2133 corresponds to 60000/2123 = 28.26 ms latency under load, while the Base RTT of 16ms corresponds to 60000/16 = 3750 RPM, son on this link load reduces the responsiveness by 3750-2123 = 1627 RPM a reduction by 100-100*2123/3750 = 43.4%, and that is with competent AQM and scheduling on the router. > > Without competent AQM/shaping I get: > ==== SUMMARY ==== > Upload capacity: 15.101 Mbps > Download capacity: 97.664 Mbps > Upload flows: 20 > Download flows: 12 > Responsiveness: Medium (427 RPM) > Base RTT: 16 > Start: 10/23/22, 13:51:50 > End: 10/23/22, 13:52:06 > OS Version: Version 12.6 (Build 21G115) > latency under load: 60000/427 = 140.52 ms > base RPM: 60000/16 = 3750 RPM > reduction RPM: 100-100*427/3750 = 88.6% > > > I understand apple's desire to have a single reported number with a single qualifier medium/high/... because in the end a link is only reliably usable if responsiveness under load stays acceptable, but with two numbers it is easier to see what one's ISP could do to help. (I guess some ISPs might already be unhappy with the single number, so this needs some diplomacy/tact) > > Regards > Sebastian > > > > *) Seemingly as quite some ISPs operate their own speedtest servers in their network and ignore customers not reaching the contracted rates into speedtest-servers located in different ASs. As the product is called internet access I a inclined to expect that my ISP maintains sufficient peering/transit capacity to reach the next tier of AS at my contracted rate (the EU legislative seems to agree, see EU directive 2015/2120). > > **) Most do by creating load themselves and measuring throughput at the same time, bounceback IIUC will focus on the latency measurement and leave the load generation optional (so offers a mode to measure responsiveness of a live network with minimal measurement traffic). @Bob, please correct me if this is wrong. > > >> >> >> On Fri, Oct 21, 2022, 5:20 PM Dave Taht <dave.taht(a)gmail.com> wrote: >> One of the best talks I've ever seen on how to measure customer >> satisfaction properly just went up after the P99 Conference. >> >> It's called Misery Metrics. >> >> After going through a deep dive as to why and how we think and act on >> percentiles, bins, and other statistical methods as to how we use the >> web and internet are *so wrong* (well worth watching and thinking >> about if you are relying on or creating network metrics today), it >> then points to the real metrics that matter to users and the ultimate >> success of an internet business: Timeouts, retries, misses, failed >> queries, angry phone calls, abandoned shopping carts and loss of >> engagement. >> >> https://www.p99conf.io/session/misery-metrics-consequences/ >> >> The ending advice was - don't aim to make a specific percentile >> acceptable, aim for an acceptable % of misery. >> >> I enjoyed the p99 conference more than any conference I've attended in years. >> >> -- >> This song goes out to all the folk that thought Stadia would work: >> https://www.linkedin.com/posts/dtaht_the-mushroom-song-activity-69813666656… >> Dave Täht CEO, TekLibre, LLC >> >> -- >> You received this message because you are subscribed to the Google Groups "discuss" group. >> To unsubscribe from this group and stop receiving emails from it, send an email to discuss+unsubscribe(a)measurementlab.net. >> To view this discussion on the web visit https://groups.google.com/a/measurementlab.net/d/msgid/discuss/CAA93jw4w27a…. >> _______________________________________________ >> Rpm mailing list >> Rpm(a)lists.bufferbloat.net >> https://lists.bufferbloat.net/listinfo/rpm > > _______________________________________________ > ippm mailing list > ippm(a)ietf.org > https://www.ietf.org/mailman/listinfo/ippm

1 0