Bandwidth utilization has always been sort of an honor code system. What I mean when I say that is that businesses seek out a service provider to provide x-amount of bandwidth, the service provider fulfills their request, and the company pays x-amount of dollars for that allotted bandwidth.
The idea seems pretty rudimentary, but the dirty secret is that more times than not, folks are paying for x-amount of bandwidth and are just not getting what they are paying for.
The reason may lie with the service itself and the service provider, but what usually occurs nine times out of ten is congestion and bottlenecks on the network; prohibiting the network from operating at its optimal performance level. No matter how you slice it, companies are not getting what they pay for--and that is just not right.
Networks are dynamic, which means they are changing constantly, so bandwidth utilization and throughput are also changing, all too frequently. In order for businesses to keep track of their bandwidth and find out if they are getting what they pay for, they need to invest in the proper network performance utility. It seems like a pretty simple solution to a very big problem, no?
However, the problem here is that the majority of solution providers choose the wrong type of networking monitoring tool for their customers. Whether it is lack of knowledge on the subject or lack of funds, the decision is made incorrectly most of the time, and it is the customer who suffers.
Many times solution providers choose a network monitoring tool that passively monitors devices on the network and simulates traffic by flooding the network. That method provides some information about the network, but it does not shed light on what is most important--how effective is throughput from one end of the network to the other?
The proper utility should send small signature packets--with the emphasis on small--through a network multiple times so the administrator can evaluate and determine the characteristics of that network. Administrators can then make the proper adjustments when problems arise.
Apparent Networks, a Vancouver-based company, makes a utility called AppareNet 2.5.1 which works almost exactly like the above scenario. It presents a series of algorithms which sends packets through a network several times to find the problematic spots. Once the packets complete their revolutions, the signatures on the packets are fully evaluated. The CRN Test Center evaluated AppareNet 2.5.1 in a recent and were really impressed with the results.
Also, in speaking with solution providers, the old adage, "If it ain't broke, don't fix it" certainly applies. If a network is operational and no complaints have been made about throughput, even though it is not operating at an optimal level, how can you fix something that you know nothing about?
Again, with the proper network monitoring utility, performance levels can constantly be monitored and often adjusted to achieve that extra performance boost. It should be the job of the network administrator to keep the network operating at a high level at all times and with the correct utility it can be. Sometimes fixing a problem can be as simple as moving a 1Mbyte file off the network, which will increase network performance.
Companies should make an effort to find out if the bandwidth they are paying for is the bandwidth they are getting. You'd be very surprised at your actual network throughput--or should I say lack thereof?
Vincent A Randazzese
Technology Editor, Digital Connect Magazine
Technical Editor, CRN
Anything you'd like to vent to Vincent? E-mail him at firstname.lastname@example.org.