Performance and reliability analysis of web server software architectures

Our increasing reliance on the information and services provided by modern Web servers mandates that these services be offered with superior performance and reliability. The architecture of a Web server has a profound impact on its performance and reliability. One of the dimensions used to characterize the architecture of a Web server is the processing model employed in the server which describes the type of process or threading model used to support a Web server operation. The main options for a processing model are process-based, thread-based or a hybrid of the process-based and the thread-based models. These options have unique advantages and disadvantages in terms of their performance and reliability tradeoffs. In this paper we propose an analysis methodology based on the Stochastic Reward Net (SRN) modeling paradigm to quantify the performance and the reliability tradeoffs in the process-based and the thread-based Web server software architectures. We demonstrate the capability of the methodology to facilitate systematic, quantitative tradeoffs using several examples.