How to Exploit the Top 10 Myths about Bandwidth & Data Transfer to Your Advantage
Author: Amy Armitage
How does one write an interesting article about bandwidth and data transfers without losing the interest of your readers within the first 20 seconds? This question has been haunting me the last few nights. Every time I sat down to get started, online poker, cable TV and Xbox kept calling my name. Really… it did.
I have a long history in tech support, so I figured this would be a sinch to write about, however since I’ve recently been focusing on marketing and development, the tech side of my brain has decided to take a time out. I had two choices. Google “Bandwidth for Dummies”, or call in the cavalry.
Yesterday I was chatting to one of our tech support supervisors asking for a definitive answer (basically so I won’t appear totally clueless). It can’t be that difficult. Can it? As I started to investigate, I realized that nobody really knew the differences.
There is a great deal of confusion regarding the differences and similarities between bandwidth and data transfer, and it’s also dependent on which context you use those terms and in which industry. I even started a thread on our forums about this to see what the general consensus was with our customers.
Is it true that if you surf backwards it brings your account usage down? Can bandwidth be recycled or reused? Is bandwidth the amount of space taken up when you stand all the band members side by side? We decided to thrash out the myths versus our take as we understand it (relating to web hosting). Really, I can make this informative and entertaining. Watch me.
The myths come from a series of support questions and forum posts we have come across over the years in the hosting industry.
Our take is provided by the incredible tech support team at LUNARPAGES.COM! (See? I was able to pitch the company without looking completely obvious).
*The responses provided are based on how we define and apply bandwidth and data transfer relating to web hosting. We do acknowledge there are other definitions/meanings – *defensive stance*
Myth: Bandwidth is not the same as file transfer or data transfer.
Our Take: Without getting into a university style lecture about analog wavelengths, digital wavelength counterparts, and the history of modems (both analog and digital), yes, bandwidth is the same as data transfer. To break it down from the view of web hosting companies, bandwidth is the amount of data that is transferred to/from a server, throughput is the amount of data that can pass through a connection per second, and latency is the time it takes to transfer data from one computer to another, including overhead in the routing process.
Myth: I have been told that bandwidth must go through a large enough opening to have a reliable & constant flow to the internet. How big are your openings?
Our Take: Bandwidth doesn’t matter in this respect. It is the throughput of the web hosting company’s servers that determine the reliability of the connection. We have 3 x 1 GigE uplinks, (3Gigabits/sec) and are at 34% capacity…and since you asked, it’s rather large.
Myth: Bandwidth defines the speed of your website
Our Take: No, referencing back to the first question, the amount of bandwidth your hosting company offers has nothing to do with the speed of your site. The primary factors impacting speed are the throughput of the server, the number of computers the data must pass through before it reaches you (hops), and the throughput of your own ISP connection. Viewing a website from a DSL connection will be much faster than viewing the same website on a dial up connection.
Myth: My ISP offers me 4GB bandwidth per month. Can I add the bandwidth you offer me and get 444 per month?
Our Take: No.
Myth: Data transfer is not measured on sending and receiving emails
Our Take: The transferring of any data from the server to your computer (and vice versa) count against your total usage for the month. This includes incoming and outgoing email.
Myth: If I get 400GB of bandwidth with my account, does that mean I have to upload 400 songs by the same *band* or can it be individual solo artists too?
Our Take: Bandwidth is not specific to music styles. Let’s break this down. If you had 400 songs, each 5 megs, you would have 2Gigs of content on your hard drive. If you have 400GB of transfer allotted to you each month that would mean that each song could be download 200 times without going over your transfer.
Myth: My account is using way too much CPU but I know I have not used my bandwidth allowance this month.
Our Take: Bandwidth usage and CPU usage are not the same. Bandwidth is the amount of data being sent to/from the server. CPU usage is the amount of exclusive time it takes the server to completely run a particular script or process. Sometimes a process can cause it as well. For example; if you’re receiving a high number of requests on static content, the httpd process can spike.
Myth: If I FTP my files in and out really quickly will the bandwidth last longer?
Our Take: The speed at which you transfer the data will not affect the total amount of data per month you are allowed to transfer, you can transfer data from your site as fast as needed any time, and it’s counted towards the monthly bottom line.
Myth: Most of the web hosts that offer unlimited or unmetered bandwidth apply no restrictions or limits. You get as much bandwidth as you want anywhere at anytime.
Our Take: Unlimited or unmetered bandwidth usually means that they do not set a pre-existing restriction on your bandwidth, or measure bandwidth overages and charge a certain amount per bandwidth unit. A number of companies however offer unlimited but in fact *some* limit the amount of data that is transferred by setting a maximum throughput limit on the account.
In summary let’s define Bandwidth and Data Transfer as we use it in the web hosting business:
Bandwidth is the measurement used to gauge the capacity a specific channel can use. All of our servers have at least 100Mbit/sec of bandwidth. That is what the customer can use if they wanted/had to.
Data Transfer is the actual amount of data transferred for a given time period. We measure data transfer in monthly increments, and in terabytes.
Bandwidth as it pertains to customers usually comes in two ways:
1) Initially most customers request 100Mbit/sec line to the internet which isn’t a problem as all of our servers are at least 100Mbits/sec, with shared servers being 1000Mbits/sec.
2) Committals: Our largest dedicated plan offers 2 terabytes of monthly transfer, which is more than enough for 99% of our customers, but occasionally, usually customers running adult sites (i.e. naked stuff), need more, so a committal is recommended.
We commit them to and average bandwidth (mbits/sec), and lessen the overage charge per gig.
Example:
A customer wants a plan that offers 2 terabytes of transfer a month, but thinks he/she will use at least 5 terabytes per month. So basically there will be a 3 TB overage. The 5 terabytes works out to be about 16 Mbits/sec. At $.75/gig in overage, this customer will owe about $2250 in transfer costs (3000 x $.75). This would be an example of a customer who would benefit from a commit. We would commit them to 5 TB of transfer a month, charge them significantly less than $2250, and lessen the overage. That way if they went over 5 TB, they would not suffer the same $.75/gig overage costs. It would equate to approx $.25/gig overage. We’d offer a customer like this a plan that states they agree to pay at least $1,200/month for the 5 Terabytes. If the customer were to use less than 5 Terabytes, they would still be charged $1,200 for the committal. If they do exceed that amount they don’t suffer the $.75/gig overage charge. Bottom line, it would cost a considerably less for the 5 TB’s than if they hadn’t signed the committal.
Hopefully this clears up some of the misconceptions about bandwidth and or data transfer. With that said, I’m sure our readers have differing opinions on this. I’d love to hear your comments/corrections on our accuracy and your take on this exhilarating and mystifying topic ;)
-
RapStar
-
DavePaxel
-
ritesh

