N-Tier Application

This is an important subject to think about when in the process of developing an application. Many different area need to be thought about when choosing on the architecture of an application, such as future development problems and complications, scalability, and performance. When you choose which architecture to actually use, you should choose on which of the three elements mentioned before that you think is the most important, as some choices that you will make can impact the other ones. For example, some choices that may increase and enhance performance, will affect scalability.

But what exactly is N-Tier architecture? The N-Tier architecture relays to the architecture of a certain application that has at least 3 layers or sections that are separate. Each section interacts with solely the section directly below, and has a very definitive function or role that it's responsible for.

But why should you use N-Tier architecture? Each separate section can be found on physically different servers with merely minor coding differences, meaning they can scale out and handle more load on a server. And, each individual section is completely hidden to other sections internally, making it capable to update a single section without modifying or recompiling another section.

This is an extremely powerful feature that is available with N-Tier architecture, as additional changes or features to a section can be accomplished without re-uploading the entire application. An example would be, by separating access code data from the business logic code. When the server's database changes, you simply need to change the code of the data access, because the logic code of the business stays put and doesn't change.

The N-Tier application typically contains three sections, named the data section, business section, and the presentation section. And you're probably wondering what each section really does, so lets go into that.

The presentation section has the duty of showing the user interface and "pushing" that interface using some business section objects and classes.

The business section has the duty of getting into the data section to get, edit, and delete data to and from the data section and transfer those results to the presentation section. This section also gets the duty of computing the data received and relayed to the presentation section.

The data section gets the duty of being the database for all the actual data. You'll often see that when in .net, it will be an access database or an SQL server, but it's not bound to just those.

Physical and Logical sections are often the ones that give people some trouble. First off, a logical section defines that the sections are different in terms of sets of classes or assemblies, but are still being hosted on the same exact server. Physical section defines that those sets of classes or assemblies are hosted on separate servers with some extra coding that handles the networking between the section. An example could be remoting and web services.

In determining to differ the sections physicall, or not doing so at all, is extra important. It honestly lies on the total load of your applications.