Re: [Ecn-sane] [tsvwg] SCE-L4S Flent "bakeoff" tests - thanks

Re: [Ecn-sane] [tsvwg] SCE-L4S Flent "bakeoff" tests - thanks

[Dave Taht]

I am looking forward very much to a rrul test on an asymmetric network configuration.

> -----Ursprüngliche Nachricht-----
> Von: tsvwg <tsvwg-bounces@ietf.org> Im Auftrag von Pete Heist
> Gesendet: Donnerstag, 12. September 2019 22:17
> An: tsvwg@ietf.org
> Betreff: [tsvwg] SCE-L4S Flent "bakeoff" tests
>
> The following repo is now available with a set of Flent tests and
> results comparing the behavior of SCE and L4S in six scenarios, five
> that were discussed at the IETF 105 hackathon and a sixth more recent
> one:
>
> https://github.com/heistp/sce-l4s-bakeoff/
>
> The main README contains the details:
> - info about the test setup
> - the scenarios and links to results obtained in our Portland data center, along with observations
> - installation instructions, for the foolhardy
> - current and future plans
>
> These tests cover the previously mentioned issue #17 (Interaction w/
> FQ AQMs), and raise a few more. I’ll gladly boil the observations down
> into a list if that helps. The plan is to iterate on these tests up
> until IETF 106, to help provide the IETF and community with more
> information regarding these two proposals.
>
> While the tsvwg mailing list can be used for discussion, if there are
> specific action items for the tests themselves (corrections, new
> tests, etc), those may also be submitted as issues on the repo
> (https://github.com/heistp/sce-l4s-bakeoff/issues), so we can keep
> track of them.
>
> Lastly, a technical detail, the currently posted run is missing its
> associated pcap files due to a botched rsync, so I’ll start another
> run and push that tomorrow. Meanwhile, the data files and plots are
> still useful without those.
>
> Regards,
> Pete Heist