Re: [Explicit-meas] [EXT] Re: Explicit Flow Measurements implementations

["Bulgarella Fabio (Guest)" <[e-mail]>]

Hi Ian, All,

here are some implementations of EFM techniques in quic-go if you want to have a look on them.


Spinbit: https://github.com/fabiobulgarella/quic-go/tree/v0.20.0-spinbit

Delaybit: https://github.com/fabiobulgarella/quic-go/tree/v0.20.0-delaybit

Loss Tbit (and old delaybit): https://github.com/fabiobulgarella/quic-go/tree/tim-ds-pl1-spin-free (note that this implementation is not updated to last quic-go tag as the others, as well as containing the implementation of the old version of the delaybit​; will be soon updated).

Loss QR (and spinbit together): https://github.com/fabiobulgarella/quic-go/tree/v0.20.0-spinbit_qr


About the observer, please consider the spindump (https://github.com/EricssonResearch/spindump) which includes logics for all the EFM techniques.


The Android implementation is quite experimental and for now closed source.


Here are some slides with some tests done in the android implementation and on linux using spindump:

https://github.com/fabiobulgarella/tim-materials/blob/main/TIMquic%20%26%20QuicGen%201.2.pdf

​

Best Regards,

Fabio and Massimo

​


________________________________
Da: Explicit-meas <[e-mail]> per conto di Ian Swett <ianswett=[e-mail]>
Inviato: martedì 23 marzo 2021 14:31
A: Nilo Massimo
Cc: [e-mail]; Dmitri Tikhonov; IETF IPPM WG ([e-mail]); Lubashev, Igor; [e-mail]; Isabelle Hamchaoui; [e-mail]; Mirja Kuehlewind; [e-mail]; Marcus Ihlar
Oggetto: Re: [Explicit-meas] [EXT] Re: Explicit Flow Measurements implementations

Great, thanks!

On Tue, Mar 23, 2021 at 7:31 AM Nilo Massimo <massimo.nilo=[e-mail]<mailto:[e-mail]>> wrote:
Hi Ian,
thank you for your reply.

As for our TIM activities, we will collect the material and in the next days we plan to share it via email in the IPPM WG.

Best Regards,
Massimo

____________________

TIM - Telecom Italia
Via G. Reiss Romoli, 274
10148 - Torino (Italy)
____________________

From: Ian Swett <ianswett=[e-mail]<mailto:[e-mail]>>
Sent: lunedì 22 marzo 2021 23:24
To: Nilo Massimo <[e-mail]<mailto:[e-mail]>>
Cc: IETF IPPM WG ([e-mail]<mailto:[e-mail]>) <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>; Dmitri Tikhonov <[e-mail]<mailto:[e-mail]>>; Lubashev, Igor <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>; Isabelle Hamchaoui <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>; Mirja Kuehlewind <[e-mail]<mailto:[e-mail]>>; [e-mail]<mailto:[e-mail]>; Marcus Ihlar <[e-mail]<mailto:[e-mail]>>
Subject: [EXT] Re: Explicit Flow Measurements implementations

Thanks for the summary.

Is there a plan to share(via email or presentation) what's been learned from those experiments with the IETF?  If they've been shared in the past and I forgot, please just add relevant links to this thread so they're easy for people to find.

Thanks, Ian

On Thu, Mar 18, 2021 at 7:57 AM Nilo Massimo <massimo.nilo=[e-mail]<mailto:[e-mail]>> wrote:
Hi All.

After the "Explicit Flow Measurements" draft presentation (https://datatracker.ietf.org/doc/html/draft-mdt-ippm-explicit-flow-measurements-01) in the IPPM WG on March 12th, it was asked by Ian Sweet and Martin Duke if there are any implementations of the EFM methodology on servers and clients.

Here is a list of some implementations made by the authors of the draft.

TIM/Telecom Italia has realized an implementation for delay measurement with:
- a client and a passive observer on mobile user devices, built in Android using Quic-Go (with gomobile binding) and spindump library.
  The observer executes delay measurements, both with Spin bit and Delay bit marking methods.
  The client app exchange QUIC traffic marked for Spin bit/Delay bit versus a server.

- a server built by mean of Quic-Go, generates marked traffic and supports delay measurements with Spin bit and Delay bit methods.

  (Draft https://datatracker.ietf.org/doc/html/draft-cnbf-ippm-user-devices-explicit-monitoring-01)
  This implementation have been tested on TIM production network, both with mobile and fixed access.

- Other implementations for loss algorithms (QR and T) have been tested in our lab with good results and we are planning to move them in production network for some services under our control.

Akamai and Orange have built a monitoring system with QL marking, specifically for loss measurements.
Their implementation has already been used in CDN servers to make EFM.
Litespeed Tech also implemented QL marking in their lsquic library.

Moreover, Ericsson is carrying out several monitoring initiatives using the SpinBit for delay EFM.

Best Regards,
Massimo and Fabio
 ____________________
TIM - Telecom Italia
Via G. Reiss Romoli, 274
10148 - Torino (Italy)
____________________


TIM - Uso Interno - Tutti i diritti riservati.

________________________________
[https://cdn.tim.ticdn.it/content/dam/flytoco-areapubblica-aemfe/tim_it/dem-assets/email_footer/Banner_mail_questa_e_tim_3012.jpg]<https://on.tim.it/banner-mail-dip>
Questo messaggio e i suoi allegati sono indirizzati esclusivamente alle persone indicate. La diffusione, copia o qualsiasi altra azione derivante dalla conoscenza di queste informazioni sono rigorosamente vietate. Qualora abbiate ricevuto questo documento per errore siete cortesemente pregati di darne immediata comunicazione al mittente e di provvedere alla sua distruzione, Grazie.

This e-mail and any attachments is confidential and may contain privileged information intended for the addressee(s) only. Dissemination, copying, printing or use by anybody else is unauthorised. If you are not the intended recipient, please delete this message and any attachments and advise the sender by return e-mail, Thanks.

Rispetta l'ambiente. Non stampare questa mail se non è necessario.