Performance benchmarking of dynamic Web technologies

Many Web sites today dynamically generate responses when user requests are received. Dynamic content creation enables features that might otherwise not be possible. One disadvantage of dynamically generating Web content is the impact on Web server performance. In this paper, we experimentally evaluate the impact of three different dynamic content technologies on Web server performance. The results show that the overheads of dynamic content generation reduce the peak response rate of a Web server by a factor of 2 to 8, depending on the workload characteristics and the specific Web technologies used.