Solved

How to calcuate the exact available internet speed

Posted on 2009-07-07
18
1,043 Views
Last Modified: 2013-12-14
Hi friends !

I have opened an internet cafe here in Kigali, Rwanda. And for this, I have subscribed Wireless Broadband Connection of 256 Kbps/128 Kbps download/upload. (Some experts here say that this speed is too low for running an internet cafe.) But, surprisingly, almost 90 % of internet cafe here in Kigali are running on same speed as this is the economic pricing of ISP.

I visited some internet cafe and I feel that they are running OK. But, basically, I am from India and I there, ISPs charge comparatively lower money for business internet connection. But here, going over 512 Kbps is too much expensive. The pricing of ISP is surprising !!!!!!

Right now, I have connected only a single computer (with public IP given by ISP) to internet and downloading Microsoft .NET Framework 2.0 (22.4 MB). The download speed is varying from 16 KB/sec to 30 KB/sec.

Please tell me how you "experts" calculate the exact available internet speed and outer interface (Public IP) and what bandwidth is required to run internet on 15 PCs.

Note: I will do Internet Connection Sharing on the Server PC where I have two NIC cards, one for public and another for internal LAN.

Regards.
frameworkdownload.JPG
0
Comment
Question by:JatinHemant
  • 7
  • 6
  • 3
  • +2
18 Comments
 
LVL 4

Assisted Solution

by:BrianMilnes
BrianMilnes earned 120 total points
ID: 24792269
Try a web based utility like http://www.broadbandspeedchecker.co.uk/
This will give an accurate assessment of true speed.
But you also need to consider other issues like contention, both external (in other owrds who else is sharing this bandwidth from the ISP) and internal (how many people will share it inside your cafe).
Brian
0
 
LVL 5

Assisted Solution

by:QEMS
QEMS earned 320 total points
ID: 24792323
I suppose the bandwidth required is dictated by expectation of your customers. In quite a few countries that would not be acceptable, but it depends what the customers will be doing and what their expectations are.

You won't be supporting many users watching youtube videos (maybe 1!) before your connection is choked up and other people can't surf the net.

I think the activity of the users is more critical than the number of users. You should be able to support quite a few people browsing email and some basic web pages but as soon as anyone starts trying to stream music/video from the internet then you're going to run into difficulties.

I would reccomend setting up a gateway server rather than using ICS and try to do as much caching there as possible. There are lots of free tools that can do this job (you could try http://www.smoothwall.org/), you won't need a really powerful computer to do this, but a large hard disk would be good so that you can cache as much data as possible.

Smoothwall is a firewall and contains other useful tools. Version 3 has some really nice bandwidth monitors and setting up caching is really easy to do and should give you a noticable performance boost for some your users (depending on what they are doing on the net).

There are other tools/products out there similar to smoothwall - I'm sure someone else can reccomend some. But I'll stick to mentioning smoothwall as I have experience of using that.
0
 

Assisted Solution

by:razorBlades
razorBlades earned 30 total points
ID: 24792332
0
 
LVL 5

Assisted Solution

by:QEMS
QEMS earned 320 total points
ID: 24792484
You could try http://speedtest.net/ and they have a server in Kigali so it should give you a better idea of speed than tests from further afield.

From the information in your post 30KB is pretty good from a 256Kbps connection, remember theres 8 bits in a byte so if you hit the maximum theoretical transfer you would be at 32KB.
0
 

Author Comment

by:JatinHemant
ID: 24797115
Thanks to all of you for your valuable comments. But please hold on and let me work on your suggestions.

Regards.
0
 

Author Comment

by:JatinHemant
ID: 24812597
Hi all !

I applied the tips and online tools that you provided. Let me tell you want I did. (Please see the snapshots of the tests.)

I used only a single PC that is connected to ISP (Broadband Wireless) with public IP address. I did not run  any other application, process or any download while testing.

Method 1. I downloaded an Audio Driver from Dell site with Interent Explorer. You can see that Data Transfer Rate is showing: 72.7 KB/sec. IT IS NOT TRUE because the data download was too slow. And according to the formula (QEMS's Comment), it results to 581.6 Kbps. That is not possible because I have subscribed only 256 Kbps/sec and for that also I am very sure that my ISP is not providing that speed.

Method 2. I downloaded the same Audio Driver from Dell site with Firefox. You can see that Data Transfer Rate is showing: 8.6 KB/sec. It results to 68.8 Kbps/sec. If It is true then I am getting very less compare to what I have subscribed.

Method 3. I performed the on-line speed tests also and the speed results are varying.


Please tell me whether the ways I used to calculate the speed (in Method 1 and Method 2) are right or not. And which test can give me the exact speed I am getting.

Also tell me, if the speed I am getting is tremendously less than 256 Kbps (that I subscribed), should I conclude that I am being given "Shared Bandwidth" !!! In that case I will complain against my ISP because before subscription they told that "256 Kbps / 128 Kbps (download/upload) will be completely for you".

Regards.
0
 

Author Comment

by:JatinHemant
ID: 24812645
Wait...sorry for not attaching snapshots. Let me attach them. They are here.
Download-Rate-in-IE.JPG
Download-Rate-in-Firefox.JPG
Download-Rate-Result-from-McAfee.JPG
Download-Rate-Result-from-Downlo.JPG
0
 
LVL 5

Accepted Solution

by:
QEMS earned 320 total points
ID: 24812768
Downloading drivers from the internet (or files in general whatever they may be) is not the best test to calculate your speed. Many things can affect the download speed such as contention on the connection (this may or may not be the case with your package) the server traffic at the location you are downloading from, network congestion anywhere on the internet, any hardware fault (yours/your ISPs/other) or whether any caching is done at your ISP.

The best way to test the speed you can get is by using one of the speedtest sites that were given. If you are unsure about the result you can try several of these sites and try to work out an average - but if they are all giving similar results you probably have a fairly accurate idea of your speed.

Try http://speedtest.net/ and initially try to use the server that is in Kigali if it is available. It was listed when I checked the site when I previously posted but I can't see it on there now. The site will reccomend the best server for you to try when you connect, but there is nothing stopping you from running several tests from different servers listed on that site.

You can also compare any results from speedtest.net with some of the other options that have been suggested in this thread such as http://www.broadbandspeedchecker.co.uk/ or http://us.mcafee.com/root/speedometer/default.asp

Some other options that you could try are:
http://www.thinkbroadband.com/speedtest.html
http://www.speakeasy.net/speedtest/
http://www.cnet.co.uk/misc/speedtest/
http://www.bandwidthplace.com/
http://www.internetfrog.com/mypc/speedtest/
http://www.dslreports.com/stest

There should be more options there than are necessary to get a very good idea of your actually connection speed.

There are software tools available where internet connection speed can be monitored, but I don't use that method so couldn't reccomend one in particular.

Or finally if you set up a smoothwall server, that has some really good bandwidth monitoring tools built in and they are very easy to use. That can even tell you (in real time) how much bandwidth is being used by each client on your network. There are other similar tools available but as I have not used them I can't reccomend any others - but I was very impressed by the features in smoothwall 3 and there are certainly some features (such as QoS) that I would definately want to have available to me if I were setting up a system like you have.

The first (and quickest and easiest) step should be to try out the speedtest sites listed and see what results you get.

Good luck with it and let us know how you get on.
0
 
LVL 5

Assisted Solution

by:QEMS
QEMS earned 320 total points
ID: 24812813
I started typing my last post before you added the screenshots but got distracted!

The McAfee speed looks low - it also says KBPS on the dial, but I think it should be Kbps

The broadband speed checker seems pretty close to the theoretical limit on your connection. Remember that you are unlikely to achieve 100% of the theoretical speed.

Try a few of the others, particularly if you can find some speed test sites that are located a bit closer to you geographically and see how you get on.
0
Maximize Your Threat Intelligence Reporting

Reporting is one of the most important and least talked about aspects of a world-class threat intelligence program. Here’s how to do it right.

 
LVL 32

Assisted Solution

by:Kamran Arshad
Kamran Arshad earned 30 total points
ID: 24821646
Hi,

Please read the PAQ for calculating the bandwidth available to you;

http://www.experts-exchange.com/Networking/Network_Management/Q_23325679.html
0
 

Author Comment

by:JatinHemant
ID: 24847399
Please hold on...
0
 
LVL 5

Assisted Solution

by:QEMS
QEMS earned 320 total points
ID: 24847530
I just tried the speedtest.net site again and the server is visible in Kigali and is provided by Rwandatel

Just for reference I did a test at the reccomended server for me and at the Kigali server. I have attached the results for reference.

Try to do an average of multiple tests from multiple sites to get the most accurate picture.
UK-Speedtest.JPG
Rwanda-Speedtest.JPG
0
 
LVL 4

Assisted Solution

by:BrianMilnes
BrianMilnes earned 120 total points
ID: 24847900
Would you explain for everyone what the relevance of the location of the speed test server is?
So we see that the latency is poor to Kigali at over ½ sec, and that might be expected. But it is also clear that that the server there is not well provisioned. So what is the point of us testing its speed?
More relevant (surely) is testing download speeds from well provisioned international sites such as Microsoft, HP, IBM and the likes who JatinHemant's customers are likely to be accessing.
Brian
speed-test-kigali.jpg
speed-test.jpg
0
 
LVL 5

Assisted Solution

by:QEMS
QEMS earned 320 total points
ID: 24848322
I assumed that as JatinHemant is in Kigali, that would be the best location for him to speedtest his server from. There are a variety of factors that could affect his performance - but just because he can't get close to the theoretical maximum speed via the well provisioned internation sites you mentioned doesn't automatically mean that he isn't getting the connection speed he is paying for.

I was surprised to see the difference in reported bandwidth between a UK server and the one in Kigali, the latency is to be expected - but bandwidth and latency are different things, even with a very high latency it should be possible to get a high throughput on the connection. It may be that these servers are inadequately provisioned, or traffic from the country that the server is located in takes priority or it may be that the "window" available for the speedtest does not take latency into account leaving less time for the actual transfer of data - these are just guesses I couldn't say for sure.

I also assumed initially that the server in Kigali was not well provisioned so I tried another server in Sydney which also showed a very low connection speed when you would expect otherwise. I have now also tried one in South Korea as the infrastructure there tends to be good and I received an equally surprising result.

These results are not what I expect, but it does show that there is a geographical discrepancy between what you would (or should?) expect and what you actually see in these results.

I can see the point about measuring the speed of the sites that customers will actually be using, but who says that they will necessarily be using MS/HP/IBM sites? They may be, they may not be, it looks like quite a big assumption to me.
Also in terms of figuring out whether or not JatinHemant is on shared bandwidth as he puts it (I'm assuming contention here - please correct me if I'm wrong), just because he may not get maximum speed from a western site or speedtest, it doesn't automatically mean that he is not getting the connection speed he is paying for.

If anyone else can offer some information about the speed discrepancy between servers in different geographical locations I would also be interested to hear their thoughts.
Sydney-Speedtest-result.JPG
Seoul-Speedtest-result.JPG
0
 
LVL 4

Assisted Solution

by:BrianMilnes
BrianMilnes earned 120 total points
ID: 24848742
Hi QEMS
You appreciate the reason for asking the question now. The Speedtest.net servers are inconsistent and many of them do not provide maximum throughput. I earlier also tested their servers in San Francisco and San Diego - I got maximum throughput from the latter and just half of that from the first.
From this we can assume that their connections are limited, either by raw connection rates, or, more likely, bandwidth throttling. The Kigali server is likely to be suffering both.
That's why I suggested using the three sites mentioned, not because they would be representative, but because they are so well provisioned. I've got 1MB/s (10Mb/s) from all of them, proving that my provider IS providing me the nominal speed.
As you say, Latency, indicated by the long Ping returns, doesn't bear any relation to the bandwidth available.
Brian
0
 

Author Comment

by:JatinHemant
ID: 24857538
Hi experts !

You are really experts, and let me first understand what tests you have done. Please give me time to read and understand what you are discussing.

Regards.
0
 

Author Comment

by:JatinHemant
ID: 24857578
Hi experts !

You are really experts, and let me first understand what tests you have done. Please give me time to read and understand what you are discussing.

Regards.
0
 

Author Comment

by:JatinHemant
ID: 25021975
Hi experts !

Thanks for being part of this discussion. Finally, I would say that I learned something good from this discussion.

Regards,

Hemant
0

Featured Post

How to run any project with ease

Manage projects of all sizes how you want. Great for personal to-do lists, project milestones, team priorities and launch plans.
- Combine task lists, docs, spreadsheets, and chat in one
- View and edit from mobile/offline
- Cut down on emails

Join & Write a Comment

Join Greg Farro and Ethan Banks from Packet Pushers (http://packetpushers.net/podcast/podcasts/pq-show-93-smart-network-monitoring-paessler-sponsored/) and Greg Ross from Paessler (https://www.paessler.com/prtg) for a discussion about smart network …
ADCs have gained traction within the last decade, largely due to increased demand for legacy load balancing appliances to handle more advanced application delivery requirements and improve application performance.
Viewers will learn how to connect to a wireless network using the network security key. They will also learn how to access the IP address and DNS server for connections that must be done manually. After setting up a router, find the network security…
Here's a very brief overview of the methods PRTG Network Monitor (https://www.paessler.com/prtg) offers for monitoring bandwidth, to help you decide which methods you´d like to investigate in more detail.  The methods are covered in more detail in o…

707 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question

Need Help in Real-Time?

Connect with top rated Experts

20 Experts available now in Live!

Get 1:1 Help Now