To Cloud Or Not To Cloud Top 5 Posts With The Most Views

In fact, the cloud is not new. As early as the 1970s, computing was commonly done in a cloud – simply because many companies, including large ones, could not afford their own mainframe computers. Instead, in what was then known as “timesharing”, businesses would “dial in” to remote computers managed by time-sharing services. The users of such services would send their own data to, and do their analysis on, the managed computers. Sound familiar?

Further, because many companies were using a shared environment, it was possible for those users to easily tap into data of common interest. The timesharing service would maintain databases of demographic and econometric data, for example, and users could incorporate that information in their analysis. This idea has not yet taken hold in the world of cloud computing, but as we will see, it is on the way.

When PCs became more prevalent, it was no longer prohibitively expensive to own computers, so the time-sharing notion fell out of fashion. But what comes around goes around, as the saying goes – and now the idea of creating hosted computer infrastructure for data storage has returned as cloud computing.

So the newness argument doesn’t really make sense anymore.

The pros and cons of using the cloud could be deemed analogous to riding public transportation. Why does someone use shared public transportation versus driving a private car? The benefit is that you don’t have to do it yourself. The time you spend on driving yourself can be spent doing something else, thereby (arguably) improving your productivity.

Now, why would one use private transportation over public? In some instances, public transportation is not flexible enough. It can be inconvenient to even get to a location where public transportation is available, as opposed to using your car.

This analogy doesn’t hold up when dealing with the cloud, because the cloud is inherently flexible. So in this respect, the argument to use the cloud is even more compelling than the argument in favor of public transportation. And yet no one questions the value of public transportation. Many, however, question whether their data warehouses should exist in the cloud.

Reduced Latency: An Argument for On-site Data Storage

There are some instances when companies might want to consider maintaining data internally. When the speed of conducting a transaction is critically important, the cloud simply cannot compete.

Take the financial services industry, and Wall Street in particular. There is a growing need for trading companies to have their computers located physically nearby a stock exchange’s computers. In high-frequency trading, trades are conducted by computers not only in milliseconds, but in microseconds.

In a business where milliseconds or less make a difference, the physical location of one’s data — as well as everything else about managing that data – is critical. It’s simply a fact of physics that it takes less time for an electrical signal to move from, say, New York to Chicago than it does from New York to London. The closer a computer is physically to the computer with which it is conducting a transaction, the less latency there will be when completing that transaction.

So, if your company has those types of operational concerns, it makes no sense to use the cloud. You want your computer as close as possible to the one with which you’re communicating and you want complete control over the infrastructure.
This is not a particularly common business problem though. While faster is usually better, for many companies that type of response time simply doesn’t matter. A few extra milliseconds is no bother for most company operations.

Security: The Concern That Isn’t

Beyond the reduced latency benefit that on-site data storage may have for some businesses, some companies feel more comfortable with their own security safeguards over those of cloud providers when it comes to sensitive information.