Summary

In this chapter we introduced the WebSphere brand of products, focusing especially on the Application Server and Network Deployment Manager products. We took some time to build up a working vocabulary, consisting of such essential terms as cell, node, server, and cluster. Combining the product names and terms, we constructed some simple WebSphere configurations to clarify the meaning of terms.

We then took a divergence into a short study of a server's anatomy. We saw that it takes a number of components to compose the internals of a J2EE server. While many of these components are easily predicted after reading a J2EE specification, a number of them are unique to the WebSphere implementation, especially the various caches and pools used to boost performance, the segmented CosNaming namespace implementation, and the JMX administrative communication backbone, that flows Admin requests through a network of app servers.

We finished the chapter up with a short study of several common WebSphere topologies. We considered the motivating factors in branch and departmental computing models that would compel us to organize our WebSphere topology differently. We also spent some time looking at larger networking configurations, including firewalls, and edge-of-network caching. We finished up our study by looking at the large-scale problems and saw how WebSphere addresses these issues through server clustering and the ability to host WebSphere applications on large zSeries platforms.

Most production configurations of WebSphere require security and administration. The next two chapters explore these areas of systems management in detail.