Vodacom Speed Test & Results: speed.vodacom.co.za & speedtest.net

Hummercellc

Expert Member
Joined
Jan 6, 2008
Messages
3,451
Speed test 1.8 & 3.6

This is weird,

when i have the 3.6Mbps trail enabled i get slower speeds

Downlink: 827Kbps
Uplink: 122Kbps

When i disable 3.6Mbps (1.8Mbps) i get higher speeds

Downlink: 1563Kbps
Uplink: 301Kbps

Seems the 3.6 don't work for me v3g?
 

vodacom3g

Vodacom Representative
Joined
Jan 14, 2005
Messages
12,065
HHHhhmmm Anyone else seen any results like this ? NOT calling you a "marketing-guru" V3G :D Just trying to see what is happening in the real world :confused::D

KZNChris :D

As I've posted before, you need to understand the network architecture and then overlay your test results on that and you'll get a better picture of what is happening.

Think of the Vodacom network as a distributed star model (which it is). Draw a picture with a star in the middle, a star on each branch and again a start on each branch. And yet again. Do this for 4 'layers' and you have a feel for the architecture. As a end user you sit on the end of one of the final branches. Think of the point where you enter the internet as the center of the middle star. The final end-points (indicating you as an end user) counts in the millions. The second layer, many thousands.

Firstly, you can see just how big this network is with thousands upon thousands of interconnecting links, all supplied by Telkom today (Not quite all, but >99%).

You can also see that as you go deeper into the core network, you more and more share links (i.e. bandwidth) with other users and most importantly, any congestion along that path will affect you, even if the congestion is not near you.

To test the speed of the Vodacom portion of the network we put speed.vodacom.co.za at the heart of the central star. The speed test is the same test system as used by other testing sites such as speedtest.net, just with a custom GUI.

So the results you get from speed.vodacom.co.za are very real (we have no way of faking them in any case) but was specifically put down so we can get a feel for the portion of the network under our control. It won't measure what you experience as you leave the Vodacom-only portion of the network.

In the distributed star model above, the second last layer represents the end points of any Telkom lines and it's here we are awaiting thousand upon thousands of lines to be delivered. Sometimes we get them quickly, mainly because Telkom had capacity available, such as the place on the N4 I posted above. Obviously there was capacity and when we placed the country-wide order for "3.6 back-haul" Telkom could deliver immediately.

But in areas where there is a much higher density, many people are asking for capacity including Vodacom and MTN. And it's here where the delays in getting bandwidth is felt the most. So you'll get 2.5Mb/s in Gat-Sonder-Water but might struggle to get 700Kb/s in Sandton. It's a pain we all would not like to feel.

Once you've tested to speed.vodacom.co.za, you know the effective throughput you get to the point before you break out on the internet. But you use very few if any servers on the Vodacom network so the info is often of (great) academic value.

To test what your experience is on the net itself you must now run other speed tests, such as the ones provided by speedtest.net. But it's not so simple, they are themselves subject to bottlenecks as seen by the Cape Town speedtest.net server in the last week or two.

But they do give an indication and if you use their results together with other test results you'll built up a picture of what happening.

At the end of the day, getting a view on your real-time usage is still the best and I use DU Meter for this. So when I download anything, I watch DU Meter and that gives me my 'real-world' experience. But it cannot tell you where any bottlenecks might be. These can be ANYWHERE between you and the remote server. The speed tests will help you narrow down where you might be seeing them.

As soon as our POP in New York comes on-line, I'll ask to a have a second Vodacom speed test installed there. This will give us great info on your experience from your PC right to where you break out in the US. Will probably do London as well.

So, keep on posting speed.vodacom.co.za results (together with other servers) as it helps us build a picture of the network.
 

vodacom3g

Vodacom Representative
Joined
Jan 14, 2005
Messages
12,065
This is weird,

when i have the 3.6Mbps trail enabled i get slower speeds

Downlink: 827Kbps
Uplink: 122Kbps

When i disable 3.6Mbps (1.8Mbps) i get higher speeds

Downlink: 1563Kbps
Uplink: 301Kbps

Seems the 3.6 don't work for me v3g?

How did you disable 3.6? :confused:
 

Hummercellc

Expert Member
Joined
Jan 6, 2008
Messages
3,451
How did you disable 3.6? :confused:

Off the v4me website. i got a sms to say it has been disabled. did a speed test then enabled it again, got another sms saying it's enabled did a speed test. then disabled it again got the sms did a speed test. and i am sure that 1.8 preforms better then 3.6 in my area.

i know its weird but just though i mention it. don't want to start a argument or anything.
 

vodacom3g

Vodacom Representative
Joined
Jan 14, 2005
Messages
12,065
Off the v4me website. i got a sms to say it has been disabled. did a speed test then enabled it again, got another sms saying it's enabled did a speed test. then disabled it again got the sms did a speed test. and i am sure that 1.8 preforms better then 3.6 in my area.

i know its weird but just though i mention it. don't want to start a argument or anything.

Hmm, I think those SMS's are probably misleading. My understanding is that whatever change you make, will only be effective the beginning of the following month. Will check up.

I think you just saw the sporadic nature of ding throughput tests in a live network. You can try and average the results out by running a number of tests.
 

AirWolf

Honorary Master
Joined
Aug 18, 2006
Messages
24,404
As I've posted before, you need to understand the network architecture and then overlay your test results on that and you'll get a better picture of what is happening.

Think of the Vodacom network as a distributed star model (which it is). Draw a picture with a star in the middle, a star on each branch and again a start on each branch. And yet again. Do this for 4 'layers' and you have a feel for the architecture. As a end user you sit on the end of one of the final branches. Think of the point where you enter the internet as the center of the middle star. The final end-points (indicating you as an end user) counts in the millions. The second layer, many thousands.

Firstly, you can see just how big this network is with thousands upon thousands of interconnecting links, all supplied by Telkom today (Not quite all, but >99%).

You can also see that as you go deeper into the core network, you more and more share links (i.e. bandwidth) with other users and most importantly, any congestion along that path will affect you, even if the congestion is not near you.

To test the speed of the Vodacom portion of the network we put speed.vodacom.co.za at the heart of the central star. The speed test is the same test system as used by other testing sites such as speedtest.net, just with a custom GUI.

So the results you get from speed.vodacom.co.za are very real (we have no way of faking them in any case) but was specifically put down so we can get a feel for the portion of the network under our control. It won't measure what you experience as you leave the Vodacom-only portion of the network.

In the distributed star model above, the second last layer represents the end points of any Telkom lines and it's here we are awaiting thousand upon thousands of lines to be delivered. Sometimes we get them quickly, mainly because Telkom had capacity available, such as the place on the N4 I posted above. Obviously there was capacity and when we placed the country-wide order for "3.6 back-haul" Telkom could deliver immediately.

But in areas where there is a much higher density, many people are asking for capacity including Vodacom and MTN. And it's here where the delays in getting bandwidth is felt the most. So you'll get 2.5Mb/s in Gat-Sonder-Water but might struggle to get 700Kb/s in Sandton. It's a pain we all would not like to feel.

Once you've tested to speed.vodacom.co.za, you know the effective throughput you get to the point before you break out on the internet. But you use very few if any servers on the Vodacom network so the info is often of (great) academic value.

To test what your experience is on the net itself you must now run other speed tests, such as the ones provided by speedtest.net. But it's not so simple, they are themselves subject to bottlenecks as seen by the Cape Town speedtest.net server in the last week or two.

But they do give an indication and if you use their results together with other test results you'll built up a picture of what happening.

At the end of the day, getting a view on your real-time usage is still the best and I use DU Meter for this. So when I download anything, I watch DU Meter and that gives me my 'real-world' experience. But it cannot tell you where any bottlenecks might be. These can be ANYWHERE between you and the remote server. The speed tests will help you narrow down where you might be seeing them.

As soon as our POP in New York comes on-line, I'll ask to a have a second Vodacom speed test installed there. This will give us great info on your experience from your PC right to where you break out in the US. Will probably do London as well.

So, keep on posting speed.vodacom.co.za results (together with other servers) as it helps us build a picture of the network.

Really interesting to know what happens behind the scenes:).
 

Big Kriss

Well-Known Member
Joined
Apr 16, 2006
Messages
481
That just confirms all our tesating and feedback re: Vodacoms Speedtest and real life down/uploads..
If your in the kernel such as Vodaworld,its blisteringly fast. Real world is that were all out on distant stars..and the distance and take off has taken its toll..

Re: the 1.8 vs 3.2 testing,in my case I swapped out a 1.8M only E620 with a 3.2 only E630
card..so I was assured of either of the 2 speeds..
Still on the former slower card..saving me bandwidth like crazy.
 

vodacom3g

Vodacom Representative
Joined
Jan 14, 2005
Messages
12,065
Still on the former slower card..saving me bandwidth like crazy.

Interesting comment, why do you find that you use more data when on the higher speed?

Do you find you browse more because things happens quicker?

Keen to hear others comments on this as well.
 

dex_10

Well-Known Member
Joined
Sep 29, 2007
Messages
222
Just Wondering

I was about to advertise my Samsung U700 cellphone on Bidorbuy to be sold. When I copied and pasted the specs of the phone from another site into the advertisement I noticed that this phone in fact is capable of 7.2 HSDPA speeds instead of normal 3.6 I stopped submitting my ad there and then!

My question is now: Should I maybe keep the phone to use as a modem instead of the Huawei E220 for once Vodacom upgrades to 7.2 or will my E220 USB modem be capable of handling 7.2?

Will be nice to have an answer from Vodacom3G. Just wondering....
 

dex_10

Well-Known Member
Joined
Sep 29, 2007
Messages
222
Using more data at higher speeds

Interesting comment, why do you find that you use more data when on the higher speed?

Do you find you browse more because things happens quicker?

Keen to hear others comments on this as well.

Very interesting indeed. I thought that it was only my imagination. Glad you asked Vodacom3G and I can give you feedback and say: Yes, I have noticed a definite increase in data usage at higher speeds, but it's not because I browse more. It just seems like I am using more data...mmm I don't know, maybe I'm wrong, but I am still doing the same things on the Net and it just seems that I am using more.

I think it will be good to hear what everybody else's opinions are.
 

Big Kriss

Well-Known Member
Joined
Apr 16, 2006
Messages
481


yea yea...ok,3.6.....:)

Im not surfing more at all...yet Im definitely using more data..I have probably 10 sites I look at daily..just to add the odd comment and see whats happening..much like here..
I allocate about 70 Megs for this each day...I found for the last 3.6 M week its been about 130 odd Megs for the same routine per day..
Im now back on 1.8 and its back to normal...
 

vodacom3g

Vodacom Representative
Joined
Jan 14, 2005
Messages
12,065
yea yea...ok,3.6.....:)

Im not surfing more at all...yet Im definitely using more data..I have probably 10 sites I look at daily..just to add the odd comment and see whats happening..much like here..
I allocate about 70 Megs for this each day...I found for the last 3.6 M week its been about 130 odd Megs for the same routine per day..
Im now back on 1.8 and its back to normal...

How is that possible? Surely if you downloaded twice the data, twice the data was requested by your PC.

I'm keen to hear where the 'extra' data is coming from? How many speed tests did you run?
 

ginggs

༼ つ ◕_◕ ༽つ
Super Moderator
Joined
Jun 26, 2006
Messages
12,151
Off the v4me website. i got a sms to say it has been disabled. did a speed test then enabled it again, got another sms saying it's enabled did a speed test. then disabled it again got the sms did a speed test. and i am sure that 1.8 preforms better then 3.6 in my area.
If you have a Huawei device you verify whether you are assigned for 1.8 or 3.6 by running the latest version of E620 Tool.
 

Darknet

Active Member
Joined
Dec 9, 2006
Messages
70
How is that possible? Surely if you downloaded twice the data, twice the data was requested by your PC.

I'm keen to hear where the 'extra' data is coming from? How many speed tests did you run?
He doesnt say that. He says he does the same thing (same routine) and it costs more.
 

AirWolf

Honorary Master
Joined
Aug 18, 2006
Messages
24,404
yea yea...ok,3.6.....:)

Im not surfing more at all...yet Im definitely using more data..I have probably 10 sites I look at daily..just to add the odd comment and see whats happening..much like here..
I allocate about 70 Megs for this each day...I found for the last 3.6 M week its been about 130 odd Megs for the same routine per day..
Im now back on 1.8 and its back to normal...

He doesnt say that. He says he does the same thing (same routine) and it costs more.

I haven't noticed any mystery data usage since Vodacom launched 3.6, my data usage patterns have not changed [except for doing a lot more speed tests than usual], and strangely enough I am not paying more than I was prior to Vodacom's launch of 3.6 on 2008-01-27.

Please post an intact quote of that entire post so we know who "he" is and what "he" actually posted.

He = Big Kriss;).

Edit: @ic why are you now a . ?:)
 

vodacom3g

Vodacom Representative
Joined
Jan 14, 2005
Messages
12,065
Increased data usage could easily be because of malware, but it would not stop when Big Kriss switch to his 1.8 device again (which is how I understand him/her).

There is now way the end device can double the downloads.

Actually Big Kriss, what you can do is to install Netlimiter (the monitor version is good enough) and then to browse with both cards again and check what's going on.

Only easy answer I can think off (outside the obvious more browsing = more data) is that on 3.6 BK did more speed tests and this added to the data count.
 

Big Kriss

Well-Known Member
Joined
Apr 16, 2006
Messages
481
Guys,sorry,I work during the day and internet in the evenings..I would have responded sooner..

Firstly..I used 3.6 for about a week..I have a 2G bundle and work it so that I have about 70Meg a day to play with.
During the week of 3.6 I was usuing about 130Meg a day.Take 20 Meg off for speed testing and downloads from Mxit and Microsoft..
I definitely noted that netmeter was showing around 130meg a day.. Now I use a Linksys wrt3gsomething router..take a further ,what,20 % off netmeter for the pc/router handshake everytime I start up ?? That shows as data transferred..

IC,I guarantee you my pc is clean..Spybot s & destroy live and scan daily,AVG subscription 4 hour scan everyday,ccleaner daily,Hijackthis daily and Microsofts spyware thingy..
Firefox with adblock,no script...No porn sites either..

Wow,where am I ? mmm ya,I said I use more data,the person that said it cost me more in money is in correct..have never said this.I control my usage very well.When Im short,I buy add hoc add ons..

As for deactivating 3.6..If you change your card to one that wont work on 3.6,I think I can safelky say Im then only downloading at 1.8M..well,in theory anyway..

V3G,Im going to do the back to back test thing with both cards..Ill post here..
Two days sufficient ? One for each?
 

vodacom3g

Vodacom Representative
Joined
Jan 14, 2005
Messages
12,065
Guys,sorry,I work during the day and internet in the evenings..I would have responded sooner..

Firstly..I used 3.6 for about a week..I have a 2G bundle and work it so that I have about 70Meg a day to play with.
During the week of 3.6 I was usuing about 130Meg a day.Take 20 Meg off for speed testing and downloads from Mxit and Microsoft..
I definitely noted that netmeter was showing around 130meg a day.. Now I use a Linksys wrt3gsomething router..take a further ,what,20 % off netmeter for the pc/router handshake everytime I start up ?? That shows as data transferred..

IC,I guarantee you my pc is clean..Spybot s & destroy live and scan daily,AVG subscription 4 hour scan everyday,ccleaner daily,Hijackthis daily and Microsofts spyware thingy..
Firefox with adblock,no script...No porn sites either..

Wow,where am I ? mmm ya,I said I use more data,the person that said it cost me more in money is in correct..have never said this.I control my usage very well.When Im short,I buy add hoc add ons..

As for deactivating 3.6..If you change your card to one that wont work on 3.6,I think I can safelky say Im then only downloading at 1.8M..well,in theory anyway..

V3G,Im going to do the back to back test thing with both cards..Ill post here..
Two days sufficient ? One for each?

Should be.

Maybe give us a slightly clearer picture on how you connect. Sounds like the 3.6 connection was directly from your PC with a E220 (?) modem? But the 1.8 card is in a router?
 

vodacom3g

Vodacom Representative
Joined
Jan 14, 2005
Messages
12,065
Should do the trick :).

Actually, when we see a 3.6 profile on a 1.8 device, we employ the new 128XQAM algorithm that employs quark-spin to transfer half the data instantaneously while the other half is done over the network. With clever reassembly you have full 3.6 capability, even on a 1.8 device.
 
Top