Mastodon Feed
Boosted by jsonstein@masto.deoan.org ("Jeff Sonstein"):
karlauerbach@sfba.social ("Karl Auerbach") wrote:
Glad to see this...
LibreQoS Bufferbloat Test
Some of my company (IWL) products can create bufferbloat situations, but testing how the end-point protocols react is a separate concern.
I've written in the past about the inadequacy of "speed test" tool and how things like "iperf" can easily be (and often are) mis-interpreted. (I'll add some links).
There's Dave Taht's piece about weakness of speed tests (I worked with him rather frequently, he added a lot to the Internet, it is sad that he is now lost to us):
https://blog.cerowrt.org/post/speedtests/
Then there is my bits about speed tests, counting bits, and iperf:
Comments on NTIA’s proposed Performance Measures for BEAD Last-Mile Networks - https://www.cavebear.com/cavebear-blog/ntia-bead-dec2024/
A Deep Dive into Bit Counting - https://www.iwl.com/blog/counting-bits
Does IPERF Tell White Lies? - https://www.iwl.com/idocs/does-iperf-tell-white-lies
Are You Being Misled by Network Speed Tests? - https://www.iwl.com/blog/why-you-shouldnt-believe-network-speed-test