[Starlink] recent starlink rrul test result

[Dave Taht]

There appears only to be variability in the baseline bandwidth of the
link on that 15sec interval. The rest is just the interactions between
our standard congestion control algorithms (cubic in this case), and a
fixed length per packet FIFO queue.

A debloated link would have 10-20ms of variability no matter the
observed bandwidth, in the third panel of this rrul plot.

The upload would have been much larger and flatter than it was in the
second panel, in part due to the fq - (and with sch_cake,
ack-filtering out much of the workload the download put on it) - see
https://blog.cerowrt.org/post/ack_filtering/

and the download bandwidth would have been about the same (across each
15s adjustment), and perhaps better as when we crack 250ms of delay in
either direction our protocols start to act up.

On Thu, Mar 3, 2022 at 9:45 AM <tom at evslin.com> wrote:
>
> Dave,
>
> How would you interpret from a bufferbloat PoV? Or is there too much other variability-induced noise to judge?
>
> Thankjs
>
> -----Original Message-----
> From: Starlink <starlink-bounces at lists.bufferbloat.net> On Behalf Of Dave Taht
> Sent: Thursday, March 3, 2022 9:36 AM
> To: Livingood, Jason <Jason_Livingood at comcast.com>
> Cc: starlink at lists.bufferbloat.net
> Subject: [Starlink] recent starlink rrul test result
>
> wildly variable bandwidth, adjusted at 15 sec intervals, induced latencies of 400ms, really terrible upload throughput relative to down.
>
> H/T David Lang for this data.
>
> On Thu, Mar 3, 2022 at 9:31 AM Livingood, Jason <Jason_Livingood at comcast.com> wrote:
> >
> > When folks experience sub-par video conferencing over Starlink, it'd be great to see working latency stats at that time - such as via the easy test at https://www.waveform.com/tools/bufferbloat.
> >
> > On 2/23/22, 09:38, "Starlink on behalf of Dave Taht" <starlink-bounces at lists.bufferbloat.net on behalf of dave.taht at gmail.com> wrote:
> >
> >
> > https://urldefense.com/v3/__https://blog.tomevslin.com/2022/02/starlin
> > ks-zoomready-rating-is-going-down.html__;!!CQl3mcHX2A!TjZjoTCGYO2gkt9x
> > YrGdFrivrejHIdNjVe0TtEsCuQfDJTyNNy4u3_5CxlN2sj8WTA26WA$
> >
> >     --
> >     I tried to build a better future, a few times:
> >
> > https://urldefense.com/v3/__https://wayforward.archive.org/?site=https
> > *3A*2F*2Fwww.icei.org__;JSUl!!CQl3mcHX2A!TjZjoTCGYO2gkt9xYrGdFrivrejHI
> > dNjVe0TtEsCuQfDJTyNNy4u3_5CxlN2sj-yqmwsLA$
> >
> >     Dave Täht CEO, TekLibre, LLC
> >     _______________________________________________
> >     Starlink mailing list
> >     Starlink at lists.bufferbloat.net
> >
> > https://urldefense.com/v3/__https://lists.bufferbloat.net/listinfo/sta
> > rlink__;!!CQl3mcHX2A!TjZjoTCGYO2gkt9xYrGdFrivrejHIdNjVe0TtEsCuQfDJTyNN
> > y4u3_5CxlN2sj_SP7yVhg$
> >
>
>
> --
> 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
>


-- 
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