Re: [Teas] AQM in draft-ietf-teas-rfc3272bis

[Dave Taht <[e-mail]>]

I certainly think that some form of aqm,fq,or fq+aqm, awareness of
correct buffer sizing for certain traffic types, notably voip, are
essential components of quality traffic engineering.

very useful e2e techniques such as packet pacing are also now widely deployed.

In lieu of actually contributing text to teas at this particular moment in time:

rfc7567 is worth citing at least.

this is a good read about short vs big buffered switches:
https://blog.cloudflare.com/asics-at-the-edge/

Over here is some recent work with graphs (if you login) on the
mikrotik deployment:

https://forum.mikrotik.com/viewtopic.php?t=179307

https://forum.mikrotik.com/viewtopic.php?t=181289

A fallout of those threads above for me is trying to discover the
actual use and deployment of wred today, particularly with or without
rfc3168 style ecn or l4s/sce style early marking.

docsis 3.1 has pie: deployment report here:
https://blog.apnic.net/2021/12/02/working-latency-the-next-qoe-frontier/
-

cisco's AFD is useful (but not an ietf standard)

fq_codel is the default in linux, ios, osx, used as a core component
in "smart queue management"
(https://www.bufferbloat.net/projects/cerowrt/wiki/Smart_Queue_Management/
) in a ton of linux based routers, and in wifi.

Middleboxes such as preseem, and libreqos are also essentially doing
what I would consider to be TE.

After dumping all that stuff at y'all, the recent bitag report might
be of some interest

https://www.bitag.org/documents/BITAG_latency_explained.pdf

and for laughs:

https://blog.apnic.net/2020/01/22/bufferbloat-may-be-solved-but-its-not-over-yet/


On Fri, Feb 25, 2022 at 4:06 AM Adrian Farrel <[e-mail]> wrote:
>
> Dave Taht (cc here) recently mentioned in a social media comment that
> draft-ietf-teas-rfc3272bis fails to mention the work of the AQM working
> group.
>
> I've invited him to suggest a couple of short paragraphs and also recommend
> where they might fit into the document.
>
> Everyone is welcome to discuss whether AQM constitutes Traffic Engineering,
> and (of course) suggest their own text.
>
> Thanks,
> Adrian
>


-- 
I tried to build a better future, a few times:
https://wayforward.archive.org/?site=https%3A%2F%2Fwww.icei.org

Dave Täht CEO, TekLibre, LLC