Speed Testing

Finding out how fast you’re connecting to the internet is a pretty simple process, but sometimes the results are a surprise. There are several sites that provide free speed tests including ours, www.pathwayz.com. The most popular around the internet are probably speedtest.net and speakeasy.net.

When you run the test, you’ll want to know how much bandwidth you should expect. That information should appear on your initial installation paperwork and on your monthly bill. Your provider should also be willing to provide you with that information when requested. You’ll have different speeds for upload (how fast you can send) and download (how fast you can receive). A T1 circuit, for example, is 1.5 Megabits per second on upload and download.

So, with the T1 in our example, you start the test on speedtest.net. The first thing you see is a PING test. A Ping is a quick test of how long a single data packet takes to hit the address of the testing machine and travel the internet back to your machine. Usually it’s a matter of milliseconds. The fewer the better, and if you test to different servers you’ll likely see that numbers vary some depending on how congested the internet is and how far away you’re sending that packet.

The next thing that will happen is the test of your throughput—your upload and download speeds. So, in our T1 example, you should be looking for something in the ballpark of 1.5 Mbps. You run the test and see—-1.3 Mbps. Wait a minute. That’s less than we’re paying for, right? Nope. Each circuit has what’s called overhead. It’s basically all that behind-the-scenes data moving back and forth keeping everything connected. So a 1.3 Mbps test is pretty good. What about 1 Mbps? There may be a problem there. So how do we know if that’s a serious issue? The best way is to see if a pattern emerges, run the test again later—maybe your internal network was slowed down because someone else was doing something intense on the circuit. Then try a different testing site and a different server. Note the details of the tests—time, site, server. Are you consistently getting that slower speed? Is a pattern emerging? That information can really help technicians troubleshoot the circuit.

I chose a T1 as our example because it’s a common low-speed circuit and because it’s upload and download speeds are the same (including the general tolerances for overhead). A circuit that has the same upload and download speed is called a synchronous circuit. Most circuits are asynchronous—their download is faster—often much faster—than their upload. So if you were running a 30 by 3 data circuit (30Mbps download and 3 Mbps upload) you’d have different results on your speed test—like, 28.4 Mbps down and 2.3Mbps up. Whether your company needs a synchronous or asynchronous circuit depends a great deal on what you do. If you place orders on the internet all day, an asynchronous circuit should be fine. If you’re uploading graphics to website, you’ll want to purchase as much upload as you can.

If you have concerns about your internet speeds, those sites can be a real benefit in finding out what you have and how it’s running. If you feel there’s been a change for the worse in speed, the tests can help determine if there’s an issue or if you need a bigger internet pipe.