Evaluations of architectural designs and implementation for database-driven web sites

Abstract

Response time is a key differentiation point among electronic commerce (e-commerce) applications. For many e-commerce applications, Web pages are created dynamically based on the current business state, stored in database systems. The architecture of database-driven e-commerce Web sites are more complex than that of typical Web sites. It requires integration of Web servers, application servers, and back-end database systems as well as synchronization of multiple databases if caches are used for acceleration of content delivery. In this paper, we analyze the factors that impact the performance and scalability of a database-driven Web site. We experimentally test (1) the performance metrics of database update, query, and synchronization; (2) trigger overhead; and (3) application server (AS) and database connection overhead and constraints. We describe several architectural design approaches for database-driven Web sites and present experimental results on their performance under various conditions, including varying request rates, update-to-request ratio, cache hit ratio, number of regional AS/data cache suites deployed, and database size. We also discuss how to handle Web page requests that involve SSL, cookies, and fragment pages (i.e. by frame or JSP).