30. Performance Tuning and Clustering

Seam drastically improves application developer productivity via extensive use of annotated POJOs, dependency bijection, and runtime service interceptors. Developers write less code because Seam generates and executes much of the boilerplate code behind the scenes. However, the developer convenience comes with a price: The more work Seam needs to do, especially at runtime, the slower the system performs. Today, as computer hardware performance continues to improve and prices continue to drop, improving developer productivity is a higher priority than raw performance.

However, for high-volume web applications, we must carefully evaluate, and try to compensate for, the performance impacts from the Seam runtime. For starters, we should tune our Seam applications to make the most of the existing hardware. If a single server is insufficient, we should also understand how to scale a Seam application by leveraging a server cluster. In this chapter, we discuss how to tune and scale Seam applications.