.

Haven't received anything as yet.
 
shack shack said:
Hi all,

It is great to see all the positive reviews of Cool Ideas, and thus is decided to sign up with them.

SADV completed physical fibre installation to and in our complex December 2017, and then we had to wait for festive season to pass.

SADV advised our trustees and body corporate that fibre is live- January 10th 2018.

I contacted Cool Ideas who, after a few checks, confirmed we have fibre in our complex.

I signed up the 10th January 2018.

On the 22 January, I logged ticket on my account to query when installation would take place- as I had heard nothing from SADV or Cool Ideas.
ticket was just closed, no answer.

25 Jan, sent an email to connect@cool ideas, to query- no response to date
That afternoon I called Cool Ideas, agents who could assist with my query were on calls, but would call me back, ref number obtained.
no call back on this to date.

26 Jan, logged a web enquiry, 10 minutes later, a man called, apologised for my poor experience, said he would query with SADV and call back to advise- still waiting for that call.

29 Jan, follow up mail sent to the mail address of the guy who called the 26th, still no response on that.

My experience so far, is far from all the great reviews, and been less than satisfying.

Starting to regret, and considering changing ISP before even getting started.

so frustrating.:mad:
Click to expand...

Just an update- to give credit where it is due.

In the time being, received an email from Cool Ideas apologising for the delay,
they are following up with SADV, and will advise as soon as SADV get back to them.
 
shack shack said:
Just an update- to give credit where it is due.

In the time being, received an email from Cool Ideas apologising for the delay,
they are following up with SADV, and will advise as soon as SADV get back to them.
Click to expand...

A further update,

Cool Ideas called me now,

SADV advise them that the complex is not live. (although SADV issued communication to this effect and advise our trustees and chairman as such)
searching various ISP providers on their sites with our address, shows fibre available- sign up.... :wtf:

So appears SADV don't know what they are doing... and incorrectly advised the ISPs that they are live- thus reflecting on all the ISP systems allowing people to sign up

We having an open day Saturday for SADV and all the ISP's to come market the fibre- most of us already signed up... :wtf:

so will take them(SADV) on then.
 
Last edited:
Have a media server in the USA that I'll be doing some tests on.

Trace to server from my router:
traceroute to 107.181.167.123 (107.181.167.x), 64 hops max, 40 byte packets
1 155.93.152.1 (155.93.152.1) 0.570 ms 0.528 ms 0.479 ms
2 154.0.3.97 (154.0.3.97) 8.138 ms 2.310 ms 2.089 ms
3 c3p-backbone.coolideas.co.za (154.0.2.133) 0.642 ms
154.0.2.1 (154.0.2.1) 0.702 ms 0.528 ms
4 154.0.4.101 (154.0.4.101) 1.227 ms 0.954 ms
154.0.3.22 (154.0.3.22) 1.368 ms
5 154.0.4.254 (154.0.4.254) 160.423 ms 160.279 ms 160.347 ms
6 v1000.core1.lon3.he.net (216.66.80.169) 159.152 ms 159.098 ms 160.760 ms
7 100ge14-1.core1.lon2.he.net (184.105.64.237) 186.173 ms 188.343 ms 159.053 ms
8 100ge13-2.core1.nyc4.he.net (72.52.92.166) 226.364 ms 226.186 ms 229.148 ms
9 100ge16-1.core1.ash1.he.net (184.105.223.165) 232.909 ms 361.993 ms 230.299 ms
10 100ge13-1.core1.lax1.he.net (184.105.80.202) 288.523 ms 288.321 ms 288.638 ms
11 100ge14-1.core1.lax2.he.net (72.52.92.122) 291.615 ms 290.414 ms 290.195 ms
12 total-server.as46562.any2ix.coresite.com (206.72.211.52) 291.518 ms 288.156 ms 287.928 ms
13 107.152.98.45 (107.152.98.45) 287.333 ms 287.057 ms 286.876 ms
14 107.152.98.102 (107.152.98.102) 288.380 ms 288.329 ms 286.975 ms
15 media (107.181.167.x) 287.279 ms 288.373 ms 288.381 ms
Click to expand...

Using speedtest-cli, I ran a speedtest to every speedtest server in South Africa and averaged the results (testing to 68 servers):
Average download: 52.89Mbps
Average upload: 28.57Mbps
Average ping: 452.83ms

Doing tests to the two CISP speedtest servers only:
Randburg:

Download: 31.59Mbps
Upload: 15.61Mbps
Ping: 352ms

Durban:
Download: 37.58Mbps
Upload 15.73Mbps
Ping: 313ms

So it seems that for some reason connectivity to CISP's network seems way below the country's average, yet the average latency is significantly higher.

Doing a direct download via http:
curl --output /dev/null http://107.181.167.x/test.iso
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 512M 100 512M 0 0 2442k 0 0:03:34 0:03:34 --:--:-- 3628k
Click to expand...

Which also falls slightly short of what would be expected given the speedtest-cli results, and significantly worse given the average results.
 
And here's a trace from the server to CISP.co.za

traceroute to cisp.co.za (154.0.13.95), 30 hops max, 60 byte packets
1 107.181.167.1 (107.181.167.1) 0.807 ms 0.773 ms 0.773 ms
2 107.152.98.101 (107.152.98.101) 0.725 ms 0.705 ms 0.714 ms
3 104.200.152.56 (104.200.152.56) 3.123 ms 3.106 ms 3.089 ms
4 hurricane-electric.as6939.any2ix.coresite.com (206.72.210.122) 0.646 ms xe-4-1.edge-ab.lax01.coloat.com (206.220.175.18) 0.596 ms 0.584 ms
5 100ge2-2.core1.lax1.he.net (72.52.92.121) 0.526 ms hurricane-electric.as6939.any2ix.coresite.com (206.72.210.122) 0.501 ms 100ge2-2.core1.lax1.he.net (72.52.92.121) 0.482 ms
6 100ge12-1.core1.ash1.he.net (184.105.80.201) 56.314 ms 56.134 ms 100ge2-2.core1.lax1.he.net (72.52.92.121) 0.307 ms
7 100ge12-1.core1.ash1.he.net (184.105.80.201) 56.095 ms 100ge1-1.core1.nyc4.he.net (184.105.223.166) 61.181 ms 100ge12-1.core1.ash1.he.net (184.105.80.201) 56.085 ms
8 100ge16-2.core1.lon2.he.net (72.52.92.165) 135.575 ms 100ge1-1.core1.nyc4.he.net (184.105.223.166) 61.130 ms 61.883 ms
9 100ge4-1.core1.lon3.he.net (184.105.64.238) 128.372 ms 128.423 ms 128.483 ms
10 cool-ideas-service-provider.10gigabitethernet5-4.core1.lon1.he.net (216.66.80.170) 128.674 ms 100ge4-1.core1.lon3.he.net (184.105.64.238) 128.868 ms cool-ideas-service-provider.10gigabitethernet5-4.core1.lon1.he.net (216.66.80.170) 128.726 ms
11 cool-ideas-service-provider.10gigabitethernet5-4.core1.lon1.he.net (216.66.80.170) 127.802 ms 128.714 ms 128.466 ms
12 c41-backbone.coolideas.co.za (154.0.1.145) 289.116 ms utx-cust.coolideas.co.za (154.0.4.53) 290.332 ms ulx-cust.coolideas.co.za (154.0.3.21) 289.688 ms
13 c35-backbone.coolideas.co.za (154.0.2.113) 288.158 ms 288.174 ms 288.816 ms
14 urm-cust.coolideas.co.za (154.0.3.226) 288.314 ms 288.379 ms 288.315 ms
15 u2n3-cust.coolideas.co.za (154.0.13.95) 288.598 ms urm-cust.coolideas.co.za (154.0.3.226) 289.497 ms 289.484 ms
Click to expand...
 
Waylander said:
Have a media server in the USA that I'll be doing some tests on.

Trace to server from my router:


Using speedtest-cli, I ran a speedtest to every speedtest server in South Africa and averaged the results (testing to 68 servers):
Average download: 52.89Mbps
Average upload: 28.57Mbps
Average ping: 452.83ms

Doing tests to the two CISP speedtest servers only:
Randburg:

Download: 31.59Mbps
Upload: 15.61Mbps
Ping: 352ms

Durban:
Download: 37.58Mbps
Upload 15.73Mbps
Ping: 313ms

So it seems that for some reason connectivity to CISP's network seems way below the country's average, yet the average latency is significantly higher.

Doing a direct download via http:


Which also falls slightly short of what would be expected given the speedtest-cli results, and significantly worse given the average results.
Click to expand...
Have you done any MTRs? Your line is performing way below what it should those latencies indicate a line fault. Are you using a VPN possibly? That would explain more of the above results.
 
Last edited:
speedtest-cli never had correct latency numbers. You can not trust them. Even from datacenter servers to speedtest servers the latency would be highly inflated. 70ms where it should be 1ms. I blame this on python itself. Someone should really write the cli over in another language which is better suited to time sensitive things. I rate go would be better.

edit: Oh wait, there is one at https://github.com/showwin/speedtest-go

Try that instead, should fix the latency numbers.
 
Last edited:
Tinuva said:
speedtest-cli never had correct latency numbers. You can not trust them. Even from datacenter servers to speedtest servers the latency would be highly inflated. 70ms where it should be 1ms. I blame this on python itself. Someone should really write the cli over in another language which is better suited to time sensitive things. I rate go would be better.

edit: Oh wait, there is one at https://github.com/showwin/speedtest-go

Try that instead, should fix the latency numbers.
Click to expand...
Also check your file size defined on your test, the default will complete before the line reaches anywhere near capacity. What happens with a UI Speedtest?
 
PBCool said:
Have you done any MTRs? Your line is performing way below what it should those latencies indicate a line fault. Are you using a VPN possibly? That would explain more of the above results.
Click to expand...

I think you misunderstand, the tests were from my server to all JHB servers as an average baseline, then tested from my server to CISP's servers for comparison. So the low performance isn't to do with my home line, but rather my server to CISP in general.
 
Tinuva said:
speedtest-cli never had correct latency numbers. You can not trust them. Even from datacenter servers to speedtest servers the latency would be highly inflated. 70ms where it should be 1ms. I blame this on python itself. Someone should really write the cli over in another language which is better suited to time sensitive things. I rate go would be better.

edit: Oh wait, there is one at https://github.com/showwin/speedtest-go

Try that instead, should fix the latency numbers.
Click to expand...

Shot
 
Using speedtest-go

beep@bop:~# speedtest-go --server 6591
Testing From IP: 107.181.167.x (Voxility Networks) [34.0494, -118.2641]

Target Server: [6591] 16691.80km Randburg (South Africa) by Cool Ideas
Latency: 341.090818ms
Download Test: ........
Upload Test: ........

Download: 17.50 Mbit/s
Upload: 42.95 Mbit/s
Click to expand...

beep@bop:~# speedtest-go --server 12018
Testing From IP: 107.181.167.x (Voxility Networks) [34.0494, -118.2641]

Target Server: [12018] 17110.96km Durban (South Africa) by Cool Ideas
Latency: 312.016544ms
Download Test: ....
Upload Test: ........

Download: 7.12 Mbit/s
Upload: 41.15 Mbit/s
Click to expand...
 
PBCool said:
Also check your file size defined on your test, the default will complete before the line reaches anywhere near capacity. What happens with a UI Speedtest?
Click to expand...

Directly downloaded a 30MB file from http://sp1.cisp.co.za/speedtest/

beep@bop:~# curl --output /dev/null http://sp1.cisp.co.za/speedtest/random4000x4000.jpg
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 30.1M 100 30.1M 0 0 2150k 0 0:00:14 0:00:14 --:--:-- 3144k
Click to expand...
 
Waylander said:
I think you misunderstand, the tests were from my server to all JHB servers as an average baseline, then tested from my server to CISP's servers for comparison. So the low performance isn't to do with my home line, but rather my server to CISP in general.
Click to expand...

Ohhk, I dont see how that is completely relative, IE your results to Cape Town will be naturally better due to the lower latency, so doing an average in that case depends on how many servers are in Cape Town etc etc. So better to rather do a test average to Johannesburg servers only.

Also your Durban latency is better than to Randburg which is basically impossible because our KZN network runs via our Johannesburg network, and that natively adds 10ms due to the longhaul portion. So you might want to use something else to give you results as mentioned by Tinuva :).
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X