Cool Ideas Fibre ISP - Feedback

Status
Not open for further replies.
Speedtest for Android just updated and also now have multi vs single connection settings (with different color "rings".

 Screenshot_20181223-211200_Speedtest.jpgScreenshot_20181223-211206_Speedtest.jpgScreenshot_20181223-211200_Speedtest.jpgScreenshot_20181223-211206_Speedtest.jpg
 
Result using Mikrotik Router: IMPECCABLE!
I am getting the full 100/25

So in conclusion it appears I am right about the Ethernet mismatch!
 
Tinuva said:
Great choice for a router
Click to expand...
Thank you. Based on the fact that the internet is now the best I've ever seen it- I am drilling holes in the wall in the next few mins to mount the Mikrotik

@PBCool please let accounts know I won't be cancelling i.e. make sure they see my ticket
 
I set out from the beginning with a RB750-GR3 and a Ubiq AC-LR. Haven't had hardware issues. Vumatel are just shockingly useless. ADSL and LTE has better uptime.
 
Mikrotik is all set up ( well almost, need to add firewall rules... will do later)
And you can see it’s working like a real FTTH connection...
Everything is freaking fast.
 
Tinuva said:
If you test to London, try pick the Coreix server.




Testing from my Ubuntu desktop pc (wired) with the following linux kernel settings:

Code: 
~$ cat /etc/sysctl.d/10-custom-kernel-bbr.conf
net.core.default_qdisc=fq
net.ipv4.tcp_congestion_control=bbr
Click to expand...


Coreix makes hardly any difference for me:

Multi thread:

1545609555044.png

Single :

1545609633994.png

Still well below what I should be getting.

All testing done on a windows 10 system.
 
any idea whats wrong in Kyalami AH?
i can't stream music videos without some buffering at the beginning of each song.
never had this issue before
 
jasweb said:
I also don't try to mix Fast Ether (100Mbps) and Gigabit (1000Mbps) networks.
Click to expand...

Exactly. If you try to put 1000Mbps of data through a 100Mbps connection you'll have to drop 90% of the packets, causing 90% of the packets to be retransmitted, causing even more slow downs, there's no way around it. Ethernet devices can only buffer a few milliseconds worth of data, if they buffer at all.
 
@PBCool - noticed my NNTP traffic was slower than usual last few days. Ran a MTR and if I'm not mistaken the traffic is being routed through Brazil (SACS cable?) even though the server is in the Netherlands?

Code: 
|------------------------------------------------------------------------------------------|
|                                      WinMTR statistics                                   |
|                       Host              -   %  | Sent | Recv | Best | Avrg | Wrst | Last |
|------------------------------------------------|------|------|------|------|------|------|
|                         router.asus.com -    0 |   95 |   95 |    0 |    0 |    0 |    0 |
|                             154.0.0.245 -    0 |   95 |   95 |    0 |    0 |    1 |    0 |
|                             154.0.4.161 -    0 |   95 |   95 |    0 |    0 |    1 |    0 |
|                             154.0.4.162 -    0 |   95 |   95 |    0 |    0 |    1 |    0 |
|                           102.130.64.13 -    0 |   95 |   95 |    0 |    0 |    7 |    0 |
|pe2-nc028.ang.sgn.as37468.angolacables.ao -   0 |   95 |   95 |   35 |   35 |   54 |   35 |
|                         170.238.232.149 -    0 |   95 |   95 |   95 |   96 |  115 |   96 |
|pe1-nc013.br.ftz.as37468.angolacables.ao -    0 |   95 |   95 |   95 |   96 |  109 |   98 |
|pe2-nce021.br.spa.as37468.angolacables.ao -   0 |   95 |   95 |  139 |  140 |  171 |  140 |
|pe1-nce014.br.spa.as37468.angolacables.ao -   0 |   95 |   95 |  139 |  139 |  146 |  140 |
|               as12989.saopaulo.sp.ix.br -    0 |   95 |   95 |  227 |  228 |  257 |  231 |
|                 xe-1-0-0.r3.ny.hwng.net -    0 |   95 |   95 |  236 |  236 |  238 |  236 |
|                      6-2.r2.ny.hwng.net -    0 |   95 |   95 |  236 |  236 |  248 |  236 |
|                        unknown.hwng.net -    0 |   95 |   95 |  245 |  245 |  251 |  245 |
|                     e3-4.r2.dc.hwng.net -    2 |   92 |   91 |  251 |  252 |  281 |  252 |
|                        unknown.hwng.net -    2 |   92 |   91 |  251 |  251 |  257 |  252 |
|                      1-1.r2.am.hwng.net -    2 |   92 |   91 |  251 |  262 |  301 |  251 |
|                           81.171.92.196 -    2 |   92 |   91 |  251 |  251 |  253 |  252 |
|________________________________________________|______|______|______|______|______|______|

Changing to the US server, I can't achieve line speed either, even with the decreased latency and 60 simultaneous threads.
 
Speed tests still look good, so if the routing is correct, perhaps I should test with a difference Usenet provider.



 
So guys, I am getting full advertised line speeds. I had to set up a few little things on the Mikrotik but that's all done now. I am truly happy today
 
ArtyLoop said:
So guys, I am getting full advertised line speeds. I had to set up a few little things on the Mikrotik but that's all done now. I am truly happy today
Click to expand...
Good stuff...
Time to do the Wingnut test...

The Wingnut Test:
1. One attempt at everything.

http://www.speedtest.net/
2. Local speedtest - auto select
3. London, first on the list.
4. New York, first on the list.

https://www.bandwidthplace.com/speedtest/
5. Bandwidth place test.

https://testmy.net/download
6. Testmy.net, manual test, 100MB.

http://www.dslreports.com/speedtest
7. Dslreports.

http://speedtest.googlefiber.net/
8. Googlefiber speedtest
 
sand_man said:
Good stuff...
Time to do the Wingnut test...

The Wingnut Test:
1. One attempt at everything.

http://www.speedtest.net/
2. Local speedtest - auto select
3. London, first on the list.
4. New York, first on the list.

https://www.bandwidthplace.com/speedtest/
5. Bandwidth place test.

https://testmy.net/download
6. Testmy.net, manual test, 100MB.

http://www.dslreports.com/speedtest
7. Dslreports.

http://speedtest.googlefiber.net/
8. Googlefiber speedtest
Click to expand...
I think we can skip five and just do single thread speedtests., as we were testing single thread.
 
sand_man said:
Good stuff...
Time to do the Wingnut test...

The Wingnut Test:
1. One attempt at everything.

http://www.speedtest.net/
2. Local speedtest - auto select
3. London, first on the list.
4. New York, first on the list.

https://www.bandwidthplace.com/speedtest/
5. Bandwidth place test.

https://testmy.net/download
6. Testmy.net, manual test, 100MB.

http://www.dslreports.com/speedtest
7. Dslreports.

http://speedtest.googlefiber.net/
8. Googlefiber speedtest
Click to expand...
Thanks I will do shortly
 
Bugger it. I can already see my download sources are flying. Screw the speed tests. That's enough proof for me.
Again, I cannot thank @PBCool enough for making me do tests that, although on the face of it were kind of redundant, but made me get a clue as to where the problem lay!
:)
 
ArtyLoop said:
Bugger it. I can already see my download sources are flying. Screw the speed tests. That's enough proof for me.
Again, I cannot thank @PBCool enough for making me do tests that, although on the face of it were kind of redundant, but made me get a clue as to where the problem lay!
:)
Click to expand...
Glad its sorted, and has also helped us possibly identify another issue with the C20s
 
Status
Not open for further replies.
Back
Top
Sign up to the MyBroadband newsletter
X