基于反模式的中间件应用系统性能优化

Anti-Pattern Based Performance Optimization for Middleware Applications

提出了一种基于反模式的中间件应用系统的性能优化方法.该方法的主要特点包括:建立了反模式元模型以使得反模式的表示更加准确:综合利用系统的静态信息和动态信息以检测运行系统中存在的反模式;系统重构在不中断应用系统运行的前提下在线执行,并在中间件的辅助下自动/半自动地完成.在J2EE平台上实现了支持该方法的原型系统,并通过网上书店的实例展示了该方法的有效性.

This paper presents an approach to optimizing performance of middleware applications based on anti-pattern. This approach has three major features： First, a meta-model is offered to build more understandable and formalized representation of anti-patterns; second, the detection of anti-patterns is based on both the static and the dynamic information, which is retrieved at rnntime; third, refactorings operate without interrupt the running systems, and is completed in an automated way with the help of the middleware. A prototype based on J2EE has been developed and an e-bookstore is used as a running example to illustrate the ideas introduced in this approach.