Mining Web Server Logs for Creating Workload Models

A3 Book section, Chapters in research books

Internal Authors/Editors

Publication Details

List of Authors: Fredrik Abbors, Dragos Truscan, Tanwir Ahmad
Editors: Holzinger A, Cardoso J, Cordeiro J, Libourel T, Maciaszek LA, Sinderen M
Publisher: Springer VS
Publication year: 2015
Publisher: Springer
Book title: Software Technologies - 9th International Joint Conference, ICSOFT 2014, Vienna, Austria, August 29-31, 2014, Revised Selected Papers
Title of series: Communications in Computer and Information Science
Volume number: 555
Start page: 131
End page: 150
ISBN: 978-3-319-25578-1
eISBN: 978-3-319-25579-8
ISSN: 1865-0929


We present a tool-supported approach where we used data mining techniques for automatically inferring workload models from historical web access log data. The workload models are represented as Probabilistic Timed Automata (PTA) and describe how users interact with the system. Via their stochastic nature, PTAs have more advantages over traditional approaches which simply playback scripted or pre-recorded traces: they are easier to create and maintain and achieve higher coverage of the tested application. The purpose of these models is to mimic real-user behavior as closely as possible when generating load. To show the validity and applicability of our proposed approach, we present a few experiments. The results show, that the workload models automatically derived from web server logs are able to generate similar load with the one applied by real-users on the system and that they can be used as the starting point for performance testing process.


Last updated on 2020-01-06 at 02:42