wingnut771

Honorary Master
Joined
Feb 15, 2011
Messages
11,189
Quite a mixed bag...
I used IDM but connections set to 1 so effectively single threaded download in every example....
I generally don't use Real-debrid for streaming but I have tested it on the new account and they worked well.
:thumbsup:
 

wingnut771

Honorary Master
Joined
Feb 15, 2011
Messages
11,189
Quite a mixed bag...
I used IDM but connections set to 1 so effectively single threaded download in every example....
I generally don't use Real-debrid for streaming but I have tested it on the new account and they worked well.
Average?
 

wingnut771

Honorary Master
Joined
Feb 15, 2011
Messages
11,189
Streaming? No idea. It peaked at around 5MB/s when it needed to. I battle to find 4k content. Maybe it's a setting I'm missing but I only ever get 1080p content.
The links I sent that you used 1 thread in IDM. If you run it for a minute what was the average speed?
 

sand_man

Honorary Master
Joined
Jun 4, 2005
Messages
31,460
The links I sent that you used 1 thread in IDM. If you run it for a minute what was the average speed?
The speeds I've posted are the average. That's exactly how I calculate. Let them run for a minute or so and guess an average speed. Obviously the smaller 100MB files run to conclusion. And I was reluctant to let the ones running at 90MB/s run to long as it seems a bit of a waste. They maxed out and maintained though, for the duration of the test.

That running for a minute would pull close to 3gigs.
 

cavedog

Honorary Master
Joined
Oct 19, 2007
Messages
18,574
@SandMan
Could you run these download tests using normal chrome download:

"Dear,

?Our servers are based in Europe and we think the speed is not good between you and our servers actually.
Your internet connection have good speed in your country but maybe not to the Europe.

Could you please send us these informations, we have to know your speed to our servers :

# Europe
- Download speed on : http://30.download.real-debrid.com/speedtest/test.rar
- Download speed on : http://proof.ovh.net/files/1Gb.dat
- Download speed on : http://5.download.real-debrid.com/speedtest/test.rar
- Download speed on : http://ping.online.net/1000Mo.dat
- Download speed on : http://mirror.nforce.com/pub/speedtests/10000mb.bin
- Download speed on : http://32.download.real-debrid.com/speedtest/test.rar
- Download speed on : http://ams.download.10gbps.io/10000mb.bin
- Download speed on : http://4.download.real-debrid.com/speedtest/test.rar
- Download speed on : http://speedtest.ams03.softlayer.com/downloads/test100.zip
- Download speed on : http://speedtest-ams3.digitalocean.com/100mb.test

For each links :

1 - Start the download
2 - Read the download speed
3 - Write the download speed
4 - Stop the download

Also, please provide us the results in a text format. Example :

- http://30.download.real-debrid.com/speedtest/test.rar -> 25 Mbit/s

Regards, Real-Debrid support team."


All of those links give me about 12-19MB/s in my browser as a single thread with Supersonic which is not bad.

real-debrid.png

@sand_man I don't think you should be happy with anything less than 10MB/s single thread to EU.
 

sand_man

Honorary Master
Joined
Jun 4, 2005
Messages
31,460
All of those links give me about 12-19MB/s in my browser as a single thread with Supersonic which is not bad.

View attachment 613224

@sand_man I don't think you should be happy with anything less than 10MB/s single thread to EU.
Agreed, on the whole they're a fairly woeful set of results for a 1Gbps line...

I'd send the results to Home-connect if, for no other reason, to prove the point that international is under performing but the use of seed boxes is widely frowned upon in the industry.... so probably prudent to shut up...
 

sand_man

Honorary Master
Joined
Jun 4, 2005
Messages
31,460

wingnut771

Honorary Master
Joined
Feb 15, 2011
Messages
11,189

sand_man

Honorary Master
Joined
Jun 4, 2005
Messages
31,460
Isn't this caused by the 95% packet loss?
Who knows? I've brought up the iperf issue and they don't seem to have any explanation for it. Their only suggestion was to ask me to use the latest version of iperf, 3.1.5 beta. They claim there's a stacking fix in that version but tests were the same on 3.1.3 vs 3.1.5. I sent them the results... Their explanation would accurately explain what I am experiencing (see below)...

iperf 3.1.3.png iperf 3.1.5.png iperf fix.PNG
 

wingnut771

Honorary Master
Joined
Feb 15, 2011
Messages
11,189

sand_man

Honorary Master
Joined
Jun 4, 2005
Messages
31,460
I would downgrade to 10MB then.
EDIT: Maybe that would wake up Vumatel.
Vumatel is without a doubt the weakest link in the supply chain! They the new Helkom, without a shadow of doubt!! I'd rather go with Openserve...
 

wingnut771

Honorary Master
Joined
Feb 15, 2011
Messages
11,189
Vumatel is without a doubt the weakest link in the supply chain! They the new Helkom, without a shadow of doubt!! I'd rather go with Openserve...
Actually 20mb adsl would be faster than 10mb fibre. That's what I used to have. In that case I'm with you, I would go back to openserve adsl. The joys of competition. The ball is in Vuma's court.
 

wingnut771

Honorary Master
Joined
Feb 15, 2011
Messages
11,189
Vumatel is without a doubt the weakest link in the supply chain! They the new Helkom, without a shadow of doubt!! I'd rather go with Openserve...
maybe @websquadza can comment why iperf has 0% loss at 8M but anything faster causes loss which indicates a config issue where something somewhere in Vuma is set to 10M not 1000M.
 

sand_man

Honorary Master
Joined
Jun 4, 2005
Messages
31,460
maybe @websquadza can comment why iperf has 0% loss at 8M but anything faster causes loss which indicates a config issue where something somewhere in Vuma is set to 10M not 1000M.
I sent iperf results on the old ONT too. All parties have all the information required to make a diagnosis IMHO. All that's missing is the political will and technical expertise, 2 crucial components....
 
Top