Benchmarking Web Server Architectures: A Simulation Study on Micro Performance

As Internet expands, the number of application servers, especially Web servers, has been increasing exponentially. To improve the performance of Web servers, researchers have paid attention to and studied the Web server’s macro-performance, namely, the response time and throughput, which can be perceived by end users directly. In this paper, we have produced a micro benchmark, ServBench, by studying the micro performance of the most widely used Apache Web server. The bottleneck functions are identified by profiling the Apache server running with a realistic workload. We select some of these functions as micro-benchmark programs and study their characteristics. We port the microbenchmark to SimpleScalar simulation environment. We obtain execution time, branch prediction and cache miss results for the microbenchmark as a function of various architectural parameters.