RWIN scaling can cause that

In reply to a post by MrSaffron:

Have tried that and I am seeing different behaviour that supports the peak speed theory

Here is a video I made of a speed test where I began a multi-thread download shortly after the test began: https://streamable.com/gkqhy

This clearly disproves the peak speed result theory, since the end result given is well below the peak speed.

A bit more fiddling and the word peak was the wrong one to use, it looks like it is using a very late sample towards the end of the test. By stopping and starting other downloads I can lower or increase the result based on the speed seen in the last sample or two it displays on screen.

Edited by MrSaffron (Tue 07-Jan-20 15:01:37)

In reply to a post by MrSaffron:

A bit more fiddling and the word peak was the wrong one to use, it looks like it is using a very late sample towards the end of the test.

I can't see much to support that either.

https://streamable.com/dehtz

In this video I started off with a multi-threaded download which I halted half way through. The result ramps up gradually after that, not suddenly, indicating the speed needle is displaying an average. The end result is well short of my 37 meg line speed.

So the needle is showing the mean at any point in time?

In reply to a post by MrSaffron:

So the needle is showing the mean at any point in time?

It looks like the mean value to date, in real-time, from the start of the test.

Thanks, sorry I started such an active debate !

I reran the test now, and it's back to what I'd expect, if not a bit better than any previous test ever:



I don't think *I* changed anything my end... now for the next challenge, contract renewal and trying to get a good deal!

Nothing changed our end either.

Does a speed drop from 67mbps to 41mbps make a great difference to "general internet use"?

I use a 14mbps connection and have no issues with HD streaming to the TV (over wifi) and 3 other devices connected concurrently.

Information was a bit vague.