WebSphere Application Server in Plain English

Let's start by saying that using WebSphere Application Server to serve content is not complicated. It just appears that way because WebSphere Application Server is meant to scale to infinity and beyond. In reality, using WebSphere Application Server to serve up content is almost as simple as A, B, C. Let's pretend we are looking a a web site for a large, complicated company. That company would have some content on the web site that would never change (maybe the company address) and some content that would be different every time you view it (maybe the status of your most recent order). Let's break a corporate web site into four parts.

A	B	C		D
The Front Door	Static Content	Dynamic Content		Back End Stuff
This is the part of the web site that faces the Internet. This is what customers think of when they say "web site". We engineers expect to find at least one router or gateway computer here.	As you might expect, static content is any web page that doesn't change. This page that you are reading right now is static content. There are servers that can spit out pages of static content really quickly. The Apache Web Server is one such server. (That is why so many web sites run on the Apache web server!) We engineers like to call computer programs that spit out static content -- like the Apache Web Server does -- a web server.	Dynamic content could be a little different each time you view it. The available balance in your checking account changes each time somebody cashes a check that you have written. The page on your bank's web site that reports your bank account balance would be a little different every time your bank processes a transaction involving your account. There are a lot of different ways to generate dynamic content. If you have a web site that is very popular, like an airline web site or a bank, you might want to run one server that spits out static content and a different server to generate dynamic content. We engineers like to call computer programs that spit out dynamic content application servers		Sometimes, what you display on your web site is a report generated by some other program. If you own a manufacturing company, you might own a program like SAP. SAP is a category of program called supply chain management software. SAP is one of several companies that create this type of software. If you want to display a list of your current suppliers, you might find it easy to ask SAP to generate that report for your web site. In this case, SAP would be an example of "back end" software. The name comes from the fact that SAP software sits behind the web site generating reports.

So it really is a little more complicated than A, B, C, but not that much more complicated!

Now let's imagine that each of those four parts of your web site are on four different computers. So now you have a small network of computers powering your web site. WebSphere Application Server can handle this.

Now let's pretend that your web site becomes so popular that you start getting a lot of visitors. Maybe you decide that you need three computers running in parallel just to keep your front door open (Column A). Maybe you find out you can get by with two computers running in parallel serving up static content (Column B). But maybe you need twelve computers running in parallel to keep up with the demand for dynamic content (Column C). All of a sudden, you web site is a lot more complicated.

Traffic can enter your web site from any of three computers. Each of those computers has to talk to one of the two computers serving up static content. Each computer serving static content might need to talk to one of twelve computers serving up the dynamic content. Each computer serving dynamic content might need to talk to one or more back end systems. All of a sudden, you have 72 separate data paths through your web site without considering the back end systems!

The moment this happens, you will find you need information about the health of each computer in your web site network and you need that information in a very timely fashion. You now have a network of data information for your customers and a network of status information for you. This article describes how WebSphere Application Server can handle this challenge. Very few other servers can.

But we are not done...

Let's pretend that you have to guarantee that your web site never fails. Ever! One way to do that is to take the computers we just described and make another site just like them. Put those new computers in a building at least 500 miles (800 kilometers) away from your company, and you are all set. Now you have a really complicated web site! WebSphere Application Server can easily monitor this kind of network. Very few other servers can. This article describes in detail how WebSphere Application Server controls such a network.

Arthur Kevin McGrath

Date:

January 12, 2007

Bio:

The author is an engineer with the consulting firm, Contract Engineers. He has consulted and lectured extensively since 1987 about the infrastructure that makes electronic commerce possible

Sources