The Nature Of Data Centers

Data centers and the nature of what they handle

data serverOne of the things that people often forget about the internet is that it’s not actually contained within the device they’re using. Obviously, the vast majority of people are well aware of that fact on an intellectual level. But it’s more the emotional view of the internet that causes a lot of confusion. When someone loads up a web page on their smartphone it feels like the machine is doing all of it by itself. Someone might have a cursory thought about the fact that the phone is making a connection to a transmission point, and from there leading to a machine which houses the information being retrieved. But people are seldom aware of just how intertwined various processes are when an url is entered into a browser.

The days when an http request would result in nothing but pure html is long gone. These days a website will usually carry multiple different types of code. Some of this code, like javascript, is handled by the browser. Or to be more accurate, it’s usually handed off to a just in time compilation engine coupled with the browser before being handed off again and rendered on the screen. Just in time compilation, or JIT, is one of the most visible processes that’s separate from HTML rendering but which will still appear on a screen. It’s rather processor intensive and one will often be able to tell when it’s occurring. A site freezing up for a second when initially rendering is often a sign of a more complex than average javascript codebase being compiled. This is also where data centers come into the picture.

Instant data centers offer a flipside to code compilation on a frontend system. The browser will probably need to work with computer code in the form of javascript. When the internet first started to gain popularity, a server would basically do as the name suggests. It would serve up data that had already been prepared. This is somewhat similar to the way a server in a restaurant receives orders and delivers the completed meal to customers. Over time the power of a data center’s servers began to grow beyond their immediate needs. This gave developers the freedom to experiment with new ideas. The end result is that servers began to not just serve up already generated content but to actually create it on the fly.

data serverToday’s modern browsers can compile javascript code. But the servers will typically work with a wide variety of scripting languages, compiled binaries, and databases. This was the first big leap which paves the way toward micro data centers. It’s a continuing pattern where the power of a server platform grows beyond the immediate needs of the tasks put in front of it. When that happens people have an opportunity to play around with the implementation of various protocols in an attempt to create something really revolutionary. With servers, the first big revolution came in the form of more dynamically generated content on the backend. This is what many people call “web 2.0”. It’s a more interactive form of the internet which works with the user in order to anticipate his or her needs.

The only problem is that for a long time this added level of processing power meant added size. The components within the industry were often far smaller than those used on a server. If a company wanted to remain competitive and offer the latest techniques than they had to keep going with large bulky models. It’s only very recently that things have changed enough to allow for the possibility of a small, or even mobile, micro data center.