Sign up with Google
Man, you are maxing out 3.6
Yes, I'm beginning to believe the Vodacom Speed Test starts collapsing past 2Mb/s. Will give more feedback soon.
-
Ask a question and get free supportUnlock new privileges and see hidden discussionsWin prizes and earn reputation
-
Question of the day ~ How often do you braai?
-
Afristay giveaway.Win a 2-night stay for two people at an establishment of your choice with R1,000 in spending money. Enter here.
Vodacom Speed Test & Results: speed.vodacom.co.za & speedtest.net
- Thread starter MrG
- Start date
Yes, I'm beginning to believe the Vodacom Speed Test starts collapsing past 2Mb/s. Will give more feedback soon.
Got the following results on the voda speed site : location JHB CBD using a E220 modem and a 64k data sim (linked to my contract sim which is 3.6 enabled):
Down:614
Up:161
5 mins later:
Down:1007
Up:115
stable hsdpa connection with 3 bars
Can vodacom give us a map of the areas where 3.6 is properly operational i.e. where ther is sufficient backhaul.
Thanks
Down:614
Up:161
5 mins later:
Down:1007
Up:115
stable hsdpa connection with 3 bars
Can vodacom give us a map of the areas where 3.6 is properly operational i.e. where ther is sufficient backhaul.
Thanks
Got the following results on the voda speed site : location JHB CBD using a E220 modem and a 64k data sim (linked to my contract sim which is 3.6 enabled):
Down:614
Up:161
5 mins later:
Down:1007
Up:115
stable hsdpa connection with 3 bars
Can vodacom give us a map of the areas where 3.6 is properly operational i.e. where ther is sufficient backhaul.
Thanks
As you know all the towers are on 3.6 but backhaul is in constant flux.
I confirmed again yesterday that there are many thousands of lines on back order. They are going in quite fast (we just seem to order faster than they can provide), so it will be near impossible to keen an accurate list.
I think the operational guys will shoot me if I add this onto their to-do lists
Is it just me or have you guys also noticed a slowdown in throughput over the last week?
Right after the launch we had many reports of ~2.5Mb/s but not anymore.
Would be good to do 3 tests:
1) speed.vodacom.co.za
2) speedtest.net to JHB
3) speedtest.net to CPT
I've been seeing better results on the off-net tests (speedtest.net) than on speed.vodacom.co.za which obviously should not be. Busy checking it out.
As you know all the towers are on 3.6 but backhaul is in constant flux.
I confirmed again yesterday that there are many thousands of lines on back order. They are going in quite fast (we just seem to order faster than they can provide), so it will be near impossible to keen an accurate list.
I think the operational guys will shoot me if I add this onto their to-do lists
Is it just me or have you guys also noticed a slowdown in throughput over the last week?
Right after the launch we had many reports of ~2.5Mb/s but not anymore.
Would be good to do 3 tests:
1) speed.vodacom.co.za
2) speedtest.net to JHB
3) speedtest.net to CPT
I've been seeing better results on the off-net tests (speedtest.net) than on speed.vodacom.co.za which obviously should not be. Busy checking it out.
Indeed.
AirWolf
Honorary Master
- Joined
- Aug 18, 2006
- Messages
- 24,404
You can say that again. London download speeds are marginally slower than Capetown for me.
Jo'burg Test
Capetown Test
London Test
Last edited:
This is a good case to discuss the effect the network architecture can have on test results and why it's not a good idea to do just one test (at one moment in time) before you accept the results.
As a starting point; it's obvious that the speed tests will download a file from a server, time this process and report it as average throughput over the download period. This is fine.
What you need to take into consideration is that you actually just measure the "slowest" element along the way. Think of a hosepipe, if you pinch it anywhere the total water flow is determined by the section that's pinched.
So anything from your PC to the Vodacom network to the non-Vodacom network (local and international internet links) to the remote server can be the bottle neck in determining maximum speed. International downloads typically shows this where the limit is the international links themselves.
It's still indicative of user experience though.
If you access the internet from the Vodacom network, you can break into the internet at a few points. These are determined by parameters like your location and APN.
So lets look at your results.
speed.vodacom.co.za was the fastest (we hope!). It sits on the Vodacom network in JHB, so if you're in CPT you still need to transverse the country-wide transmission network to get to it.
speedtest.net in CPT should give a better reading than JHB if you break out in CPT (using the internet APN in Western Cape). If it does not, it means the peering links from Cape Town to that ISP is congested or the test server itself is loaded.
If you broke out in JHB and CPT is slower, it again shows the peering links from SAIX down to that ISP is the most likely culprit.
Measuring speed in a dynamic network where load on any one of many elements can change instantaneously is nearly impossible. But taking a number of reading to different servers over time, will build up a picture. But even this 'picture' must be properly analysed before conclusions can be made. It's all part of Zen and the Art of Performance Management.
I've recently seen better result to speedtest.net than to speed.vodacom.co.za. So currently I don't fully trust the test and we're busy looking into it. I specifically think it starts under reading at around 2Mb/s.
A bit like the story of Van who got the job to test train wheels by hitting them with a hammer and listening to the sound. Cracked wheels made a dull sound and must be replaced.
Ol' Van had already decommissioned 50 trucks when they realised his hammer was cracked......
There are so many factors involved in doing these tests (links, servers, PC's, protocols, policies, etc.) that one should always carefully analyse the results. But it's very important to do these tests, so don't stop posting the results!
- Joined
- Jun 22, 2007
- Messages
- 10,896
Seeing similar issues with the speedtest.net CPT server. speedtest.net is reporting my location as CPT. This is from Oxford street just off Wellington rd out in Durbanville.
speed.vodacom.co.za
Download Speed: 2297 kbps (287.1 KB/sec transfer rate)
Upload Speed: 259 kbps (32.4 KB/sec transfer rate)
speedtest.net - JHB
down: 1908 kb/s
up: 333 kb/s
ping 182ms
speedtest.net - CPT
down: 1009 kb/s
up: 114 kb/s
ping: 158ms
speed.vodacom.co.za
Download Speed: 2297 kbps (287.1 KB/sec transfer rate)
Upload Speed: 259 kbps (32.4 KB/sec transfer rate)
speedtest.net - JHB
down: 1908 kb/s
up: 333 kb/s
ping 182ms
speedtest.net - CPT
down: 1009 kb/s
up: 114 kb/s
ping: 158ms
Which APN? Your current IP?
At least you can show ginggs the 1908 to JHB as prove that you must be in a 3.6 cell!
Internet APN?
You being in JHB but getting 1282 to the speedtest.net JHB server while morkhans got 1908 to the same server but from CPT, just highlights the points I made above about speed tests and how one should interpret them.
Notice how 'clean' the uploads are. Bodes well for UPA.
hermansteyn
Well-Known Member
- Joined
- Aug 25, 2007
- Messages
- 280
My best speed test thus far:
Download Speed: 2552 kbps (319 KB/sec transfer rate)
Upload Speed: 348 kbps (43.5 KB/sec transfer rate)
Download Speed: 2552 kbps (319 KB/sec transfer rate)
Upload Speed: 348 kbps (43.5 KB/sec transfer rate)
sand_man
Honorary Master
- Joined
- Jun 4, 2005
- Messages
- 35,844
yean static, the internet...
SUPERMAN89
Expert Member
- Joined
- Jan 14, 2008
- Messages
- 1,385
Vodacom speed in durban is great(i hope) but there's been alot of dc/cant connect/vpn apn dnt work/---besides that haha its ok
besides that haha its okmy slow speed problem found...
ive been using E220 on leopard. leopard not supported with dashboard, so we got a jippo connection running it...
i moved the E220 to my hp laptop and .... varooooooommm.... 2600kbps
now to solve the leopard issue
ive been using E220 on leopard. leopard not supported with dashboard, so we got a jippo connection running it...
i moved the E220 to my hp laptop and .... varooooooommm.... 2600kbps
now to solve the leopard issue
my slow speed problem found...
ive been using E220 on leopard. leopard not supported with dashboard, so we got a jippo connection running it...
i moved the E220 to my hp laptop and .... varooooooommm.... 2600kbps
now to solve the leopard issue
Our first MAC report, I think!
Have you tried the latest MAC dash?
Although 'officially' not supporting Leopard, we had reports here on the forum that it works fine on Leopard. Maybe try it.
Again your experience underscores the whole issue of maximising IT system performance.
Something, right from PC to the final server, can act as the bottleneck.
Best definition of Performance Management I ever heard; "PM is the detection and removal of a revolving set of bottlenecks." - Shaun Wormwald
As soon as you find the biggest bottleneck, the next one appears.