Chapter 5: WebSphere Deployment and Network Architecture

Highlights

If tuning websphere constitutes the micro view of your tuning and performance methodology, then the deployment and network architecture (aka the topologies) is the macro view. Throughout this chapter, you'll take a detailed look at the various network and deployment architectures that you can use with WebSphere.

Each topology has its place and its pros and cons, mostly dependent on or driven by three key factors: scalability, reliability (availability), and performance. In this chapter I examine each topology from a top-down perspective and focus on what each design entails, cover what its benefits and trade-offs are, and suggest different WebSphere implementations that would benefit from each of the topological architectures.

Like most topics I've discussed so far in this book, cost is the most prohibitive factor. If cost wasn't an issue, I'm sure most of us would go for the most scalable and reliable system. However, as we all know, cost benefit realization is an important analytical process that should be weighed against each design to essentially determine whether paying 50 percent more for your infrastructure costs is really worth the fourth 9 in the magic availability measurement (i.e., 99.99 percent versus 99.9 percent availability).