Hierarchical workload characterization for a busy web server

Abstract

This paper introduces the concept of a Webserver access hierarchy-a three-tier hierarchy that describes the traffic to a Web server in three levels: as aggregate traffic from multiple clients, as traffic from individual clients, and as traffic within sessions of individual clients. A detailed workload characterization study was undertaken of the Webserver access hierarchy of a busy commercial server using an access log of 80 million requests captured over seven days of observation. The behavioural characteristics that emerge from this study show different features at each level and suggest effective stategies for managing resources at busy Internet Web servers. © Springer-Verlag Berlin Heidelberg 2002.