[Cerowrt-devel] Comcast's NANOG slides re Bufferbloat posted (Oct 2016)

Dave Taht dave.taht at gmail.com
Thu Oct 20 14:29:11 EDT 2016

On Thu, Oct 20, 2016 at 11:12 AM, Mikael Abrahamsson <swmike at swm.pp.se> wrote:
> On Thu, 20 Oct 2016, Rich Brown wrote:
>> https://www.nanog.org/sites/default/files/20160922_Klatsky_First_Steps_In_v1.pdf
> Does anyone understand what access speeds these customers had during these
> tests?

What that work showed was that basically all cablemodems had a fixed
upstream buffersize that is too large by factors of 25% to a factor of
10, and that 48Kbytes was a pretty good basic sweet spot... and if
they could just fix that across what's deployed life would get better
for everybody without fancy schmancy aqms... But we knew that


I note that I was called into consult a bit on this project and don't
feel at liberty at the moment to disclose their up/down parameters or
the direction of future work. Carl's nanog talk was filmed and there
were some interesting discussion afterwards about things like BBR.


I was delighted that they used flent to exercise the connection(s).

> 96 kilobyte buffer on 1 megabit/s upstream or 50 megabit/s upstream makes a
> big difference.
> (I have 250/50 on my DOCSIS3.0 connection, but perhaps it's common knowledge
> what speeds Comcast customers typically has, that I don't know?)
> --
> Mikael Abrahamsson    email: swmike at swm.pp.se
> _______________________________________________
> Cerowrt-devel mailing list
> Cerowrt-devel at lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/cerowrt-devel

Dave Täht
Let's go make home routers and wifi faster! With better software!

More information about the Cerowrt-devel mailing list