Empirical Characterization of Session–Based Workload and Reliability for Web Servers

Abstract

The growing availability of Internet access has led to significant increase in the use of World Wide Web. If we are to design dependable Web-based systems that deal effectively with the increasing number of clients and highly variable workload, it is important to be able to describe the Web workload and errors accurately. In this paper we focus on the detailed empirical analysis of the sessionbased workload and reliability based on the data extracted from actual Web logs of eleven Web servers. First, we introduce and rigourously analyze several intrasession and inter-session metrics that collectively describe Web workload in terms of user sessions. Then, we analyze Web error characteristics and estimate the requestbased and session-based reliability of Web servers. Finally, we identify the invariants of the Web workload and reliability that apply through all data sets considered. The results presented in this paper show that session-based workload and reliability are better indicators of the users perception of the Web quality than the requestbased metrics.

Workload Modeling for Computer Systems Performance Evaluation

“…This leads to a unique reference pattern, in which all available documents are visited the same number of times, as opposed to the highly skewed distribution of popularity that characterizes human ac-cesses. As a result robot activity may distort the results of workload studies that are aimed at characterizing human behavior [297].…”

“…Few analyses of web sessions have been conducted [34,297]. Results indicate that sessions tend to be very short (few minutes and few requests), with a significant fraction containing only a single request.…”

“…An especially important aspect of sessions is their failure characteristics [297]. It is common to study failures at the level of individual requests.…”

Dependability Assessment of Web Service Orchestrations

“…[13]). The analytical models exploit different kinds of Markov processes to define availability and reliability models for a composite Web service.…”

“…The analytical models exploit different kinds of Markov processes to define availability and reliability models for a composite Web service. The empirical analyses consider both the workloads and the reliability of Web servers, and distinguish between inter-session and intra-session Web characteristics [13]. More recently, some papers have tackled the problem of composing a service-oriented system from publicly available Web services (e.g., [14]), taking into account different types of Web service failures.…”