[Rpm] Example plots for USB3 and thunderbolt Gigabit ethernet dongles

[Rpm] Example plots for USB3 and thunderbolt Gigabit ethernet dongles

[Sebastian Moeller]

[-- Attachment #1: Type: text/plain, Size: 751 bytes --]

IMHO this is a great illustration why bidirectionally saturating tests (as networkQuality does by default) are really a relevant test. I would go as far as arguing that it is better than doing upload and download independently, especially if the latency detection can converted to OWDs...

As I proposed at another place, if the latency probe would return the time from the server OWDs would be measurable (and if the OWDs are first measured without load, the deltaOWD caused by the load can even be calculated with unsynchronized clocks at both ends).

Regards
	Sebastian

P.S.: These plots are the optinal output of the crusader GUI after a successful measurement; I really like the co-registered throughput and latency graphs.


pos
  

[-- Attachment #2.1: Type: text/html, Size: 1427 bytes --]

[-- Attachment #2.2: plot 2022.06.18 23-47-59.png --]
[-- Type: image/png, Size: 112070 bytes --]

Re: [Rpm] Example plots for USB3 and thunderbolt Gigabit ethernet dongles

[Sebastian Moeller]

[-- Attachment #1: Type: text/plain, Size: 1098 bytes --]

Argh, sorry, this time with both images...

> On Jun 19, 2022, at 18:07, Sebastian Moeller via Rpm <rpm@lists.bufferbloat.net> wrote:
> 
> IMHO this is a great illustration why bidirectionally saturating tests (as networkQuality does by default) are really a relevant test. I would go as far as arguing that it is better than doing upload and download independently, especially if the latency detection can converted to OWDs...
> 
> As I proposed at another place, if the latency probe would return the time from the server OWDs would be measurable (and if the OWDs are first measured without load, the deltaOWD caused by the load can even be calculated with unsynchronized clocks at both ends).
> 
> Regards
> 	Sebastian
> 
> P.S.: These plots are the optinal output of the crusader GUI after a successful measurement; I really like the co-registered throughput and latency graphs.
> 
> 
> pos<plot 2022.06.18 23-47-59.png>


>   
> _______________________________________________
> Rpm mailing list
> Rpm@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/rpm


[-- Attachment #2.1: Type: text/html, Size: 2068 bytes --]

[-- Attachment #2.2: plot 2022.06.19 17-25-41.png --]
[-- Type: image/png, Size: 72624 bytes --]

[-- Attachment #2.3: plot 2022.06.18 23-47-59.png --]
[-- Type: image/png, Size: 112070 bytes --]