[Rpm] [Starlink] the grinch meets cloudflare's christmas present
dickroy at alum.mit.edu
Wed Jan 4 23:25:14 EST 2023
HNY to all!
Seems to me that we often get distracted by nomenclature needlessly.
Perhaps it's time to agree on the lexicon that should be used going forward
so as to avoid such distractions.
Perhaps a place to start is "the technical facts":
1) "capacity" is a property of a link (or links) that specifies the
theoretically achievable maximum error-free transmission rate of
data/information through a noisy channel (or channels, the multidimensiaonl
version of the capacity theorem). Yes, it's much more complicated than that
in general, however the basic principle is easy to understand. "You can only
get so much water through a hose of size X with an applied pressure of
2) "maximum achievable throughput/data-rate" of a channel is the maximum
rate (always <= channel capacity) at which information can be exchanged in
the channel as implemented (under all conditions).
3) achieved/measured "data rate" is the measured/estimated rate of
information transmission (always <= maximum achievable rate" for that
channel) in a channel under a given set of conditions.
4) "latency" is the amount of time it takes information to get from a
source to its destination (there may be multiple destinations each with
different latencies :-)). Latency may (or may not) include the unavoidable
consequence of the laws of physics that state information can not travel
faster than the "speed" of light (actually the "speed" in whatever medium
and by whatever mode the information is actually being transported)! Tin
cans and strings have a transmission speed that depends critically on how
hard each person at the end of the "link" are pulling on their cans! :-) The
point is that when included, information transmission times from source to
destination set a lower bound on the "latency" of that link/channel.
5) . (feel free to add more :-)
My two cents!
From: Starlink [mailto:starlink-bounces at lists.bufferbloat.net] On Behalf Of
jf--- via Starlink
Sent: Wednesday, January 4, 2023 11:20 AM
To: Dave Taht
Cc: Dave Taht via Starlink; IETF IPPM WG; libreqos; Cake List; Rpm; bloat
Subject: Re: [Starlink] [Rpm] the grinch meets cloudflare's christmas
HNY Dave and all the rest,
Great to see yet another capacity test add latency metrics to the results.
This one looks like a good start.
Results from my Windstream DOCSIS 3.1 line (3.1 on download only, up is 3.0)
Gigabit down / 35Mbps up provisioning. Using an IQrouter Pro (an i5 x86)
with Cake set for 710/31 as this ISP can't deliver reliable low-latency
unless you shave a good bit off the targets. My local loop is pretty
Here's the latest Cloudflare test:
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Rpm