why do you need to test every 40 mins though?

Jase2985:

 

why do you need to test every 40 mins though?

Options are 10-49mins. It is either have it turned off, or set at 40mins due to OCD reasons.

Jase2985:

 

why do you need to test every 40 mins though?

 

There's a few legitimate reasons, but usually it's because they can automated it and then complain if the results aren't what they expect. Sort of a way to hold their provider to account. It's a bit silly.

michaelmurfy:

Jase2985:

why do you need to test every 40 mins though?

Options are 10-49mins. It is either have it turned off, or set at 40mins due to OCD reasons.

One of the honest IPv4 speedtests I know is http://avm.de/nc/service/zack-der-speedtest-fuer-ihre-breitbandverbindung/

It's in German and from the manufacturer of the Fritzboxes but the resulting numbers are quite international :-)

michaelmurfy:

 

Jase2985:

 

why do you need to test every 40 mins though?

 

Options are 10-49mins. It is either have it turned off, or set at 40mins due to OCD reasons.

 

That's a pretty bad feature to have...

at a single user level, brilliant.

but when you look at say, a provider like 2d that i'd personally say is a tad slated towards running such gear that's killer.

Providers dont design their speedtest servers to be being hammered by hundreds of users all day long...

hio77:

 

Providers dont design their speedtest servers to be being hammered by hundreds of users all day long...

 

neither does the LFC's network or GPON :)

Jase2985:

 

hio77:

 

Providers dont design their speedtest servers to be being hammered by hundreds of users all day long...

 

 

neither does the LFC's network or GPON :)

 

fair point... few of these, with chorus's push for 'home plans' and your starting  to hit limits.

In this case appears the server before the LFC though.

It's kinda a common statement that 2d's test servers are pretty solid test points, but alot of traffic comes from that viewpoint!

I commend you for going through the efforts there!

Quite interesting to see OSX suffers as bad as windows does lately too.

Firstclass: Hi All !

OK I have hooked up the FreeBSD server to a monitor, a USB keyboard and a mouse.
I do not have a GUI installed on the FreeBSD server for obvious reasons hence the LiveUSB approach.

Booted the FreeBSD server with the Fedora version 30 Live USB and plugged it into a Netgear ProSAFE Gigabit switch that I know can do 940Mb/sec (tested it with iPerf2/3/Netperf).

Just so happens that both computers use an Intel Pro/1000 network card.

Plugged in an iMac OS-X with the latest OS updates into the same switch, so I can test them side by side,

Fedora version 30 did
868Mb/sec down and 539 up.

OS-X did
689 down and 531 up.

Conclusion, NO issues with my cabling/fibre or 2D as an ISP.
Note how the Linux TCP/IP stack destroys OS-X on downloads, but uploads are even maybe due to the 400ish limit that 2D sets.
Pretty generous overhead on upstream I think.

Just a note to those who may not know this(regular readers or pros here do) that if you enable QOS it kills your throughput. I disabled QOS for the duration of the tests, but normally I leave it enabled to deal with bufferbloat.
https://www.bufferbloat.net/projects/bloat/wiki/What_can_I_do_about_Bufferbloat/

Thanks for the tips on the LiveUSB/DVD approach and your guidance !

Firstclass: @NickMack

Thank you, yes it was, I downloaded your doco and looked at your post, much appreciated, great support !!!

I am in ChCh, using nperf.com on my mac with firefox

- chch server = 908/530 & 2ms

- akld server = 896/525 & 16ms

- dun server = 910/495 & 10ms

if you test with browser, nerf recommend firefox