The first time I heard Van talk about ECN (and, I'm sure, something like
RED), he said randomness was really important ...

Spencer

On Tuesday, July 30, 2013, Michael Welzl wrote:

> Hi,
>
> I like the slides too, but I think that slide #5 should contain:
> - counteract synchronization
>
> This is in fact a goal of AQM that I had always considered to be one of
> the key motivators behind RED (without that goal, it could probably have
> just been ED instead  :-)  ). Somewhat strangely, this goal hasn't been
> mentioned much recently.
>
> Is it just a simulation artifact? I don't think so. Yes we see it a lot in
> simulations:
> http://www.icir.org/floyd/papers/phase.pdf
> but it also appears in reality:
> http://caia.swin.edu.au/pubs/ATNAC06/Jaym.pdf
> and it would surprise me greatly if it wasn't a much more significant
> problem in more controlled environments, in particular data centers.
>
> Cheers,
> Michael
>
>
> On Jul 29, 2013, at 2:42 PM, Wesley Eddy <[email protected]<javascript:;>>
> wrote:
>
> > As BoF chairs, we were asked to talk to the RTGAREA meeting
> > at this week's IETF about the AQM BoF and what we're doing.
> >
> > The slides I'm planning to discuss there are:
> > http://www.ietf.org/proceedings/87/slides/slides-87-aqm-3.pptx
> >
> > --
> > Wes Eddy
> > MTI Systems
> > _______________________________________________
> > aqm mailing list
> > [email protected] <javascript:;>
> > https://www.ietf.org/mailman/listinfo/aqm
>
> _______________________________________________
> aqm mailing list
> [email protected] <javascript:;>
> https://www.ietf.org/mailman/listinfo/aqm
>
_______________________________________________
aqm mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/aqm

Reply via email to