Re: [aqm] adoption call: algorithm drafts

[Dave Taht <[e-mail]>]

On Tue, Sep 16, 2014 at 4:38 PM, Wesley Eddy <[e-mail]> wrote:
> On the mailing list, on the summer webex, and during the last
> meeting, we've discussed plans and expectations for adopting
> algorithm specifications as working group drafts.
>
> For a process description, see page 6 of:
> http://www.ietf.org/proceedings/90/slides/slides-90-aqm-0.pdf
> (one exception is that our charter says algorithms can be
> Proposed Standard, and we must have forgot that when drawing
> the chart)
>
> Here is the list of algorithm documents and how we view their
> current status (alphabetical order):
>
> - draft-hoeiland-joergensen-aqm-fq-codel : this has cleared
>   "Gate-1", and is welcome for continued discussion, but is
>   heavily focused on the FQ / scheduling aspect, and relies
>   on CoDel as an AQM.  It could become an appendix to the CoDel
>   draft, or proceed some other way.  Feedback about how folks
>   want to handle this is very welcome.

Hmm. We did not update this draft (2 minor nits in it) yet for a
 trivial reason: because the tracker does not handle hyphenated
names correctly.

I agree it relies heavily on the codel draft to keep the distinction
between flow queuing and aqm distinct. If it were to include
codel (or vice versa), the draft would get rather long.

As it is, it contains a textual description and not enough
pseudo-code, IMHO.

If a fq_pie were produced, how would that work?

>
> - draft-lauten-aqm-gsp : this has cleared "Gate-1" and is
>   welcome for continued discussion, but we haven't yet seen
>   the interest from multiple parties towards progressing it

I have no interest in this.

> - draft-nichols-tsvwg-codel : this is ready for the "Gate-2"
>   decision about WG adoption.  We have some significant
>   concerns about whether editors are available to handle this
>   through the working group, however.

Well, I'd be willing to take it on if the primary authors would
let me.

> - draft-pan-aqm-pie : this is ready for the "Gate-2" decision
>   about WG adoption.
>
> - draft-white-aqm-docsis-pie : this could either become an
>   appendix to the draft-pan specification or could progress
>   in parallel to it as an Informational description of what
>   has been done, and shouldn't require much additional work.

I happen to like the second draft here very much as it goes
into more of the complexities of actual real world implementation.

It didn't seem as though the working group is interested in
"comprehensive queue management" as I described in my
last preso.

(the code for that - an all-singing, all-dancing qdisc, that does
rate shaping, diffserv differentiation, AND fq_codel has been
taking shape, rapidly.)


> We would like feedback right now on adopting:
> 1 - http://tools.ietf.org/html/draft-pan-aqm-pie-01
> and
> 2 - http://tools.ietf.org/html/draft-nichols-tsvwg-codel-02
>
> towards the charter milestone for submitting algorithm
> specifications to the IESG.  Whether they are Proposed Standard
> or Experimental can be debated now or later, but we want to
> probe if there's critical mass to adopt them first.

+1 on both.

>
> --
> Wes Eddy
> MTI Systems
>
> _______________________________________________
> aqm mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/aqm



-- 
Dave Täht

https://www.bufferbloat.net/projects/make-wifi-fast