Chapter 3: Advanced Clustering Techniques for Maximizing Web Site Availability with WebSphere Application Server, Version 5

---

Overview

Executive summary: The always-on global marketplace makes continuous availability a requirement for an increasing number of Web sites. Over the last few years, IBM's High-Volume Web Sites (HVWS) team has worked on many Web infrastructure projects that required continuous availability in addition to ground breaking scalability. To address the availability requirement, the HVWS team developed several advanced techniques for clustering Web and application servers.

Many techniques for server clustering provide continuous availability as long as the consideration of other requirements is limited or removed. If scalability is not considered properly, performance bottlenecks may develop. Maintaining availability during system maintenance tasks, such as upgrades to hardware and applications, remains a challenge. With the proper techniques and the right Java 2 Enterprise Edition (J2EE) application server it is possible to achieve continuous availability and address the other requirements.

This chapter discusses advanced techniques for clustering Web and application servers to achieve continuous availability using IBM WebSphere Application Server, Version 5. Several clustering scenarios are used to illustrate the techniques. The scenarios are approached from the perspective of demonstrating continuous availability for highly scalable Web sites, including addressing system maintenance techniques. Sample scripts that use the new WebSphere Application Server administration scripting language are provided.

---