Variability in quality attributes of service-based software systems: A systematic literature review

  • Authors:

  • Sara Mahdavi-Hezavehi;Matthias Galster;Paris Avgeriou

  • Affiliations:

  • University of Groningen, Department of Mathematics and Computing Science, P.O. Box 407, 9700 AK Groningen, The Netherlands;University of Groningen, Department of Mathematics and Computing Science, P.O. Box 407, 9700 AK Groningen, The Netherlands;University of Groningen, Department of Mathematics and Computing Science, P.O. Box 407, 9700 AK Groningen, The Netherlands

  • Venue:
  • Information and Software Technology
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

Context: Variability is the ability of a software artifact (e.g., a system, component) to be adapted for a specific context, in a preplanned manner. Variability not only affects functionality, but also quality attributes (e.g., security, performance). Service-based software systems consider variability in functionality implicitly by dynamic service composition. However, variability in quality attributes of service-based systems seems insufficiently addressed in current design practices. Objective: We aim at (a) assessing methods for handling variability in quality attributes of service-based systems, (b) collecting evidence about current research that suggests implications for practice, and (c) identifying open problems and areas for improvement. Method: A systematic literature review with an automated search was conducted. The review included studies published between the year 2000 and 2011. We identified 46 relevant studies. Results: Current methods focus on a few quality attributes, in particular performance and availability. Also, most methods use formal techniques. Furthermore, current studies do not provide enough evidence for practitioners to adopt proposed approaches. So far, variability in quality attributes has mainly been studied in laboratory settings rather than in industrial environments. Conclusions: The product line domain as the domain that traditionally deals with variability has only little impact on handling variability in quality attributes. The lack of tool support, the lack of practical research and evidence for the applicability of approaches to handle variability are obstacles for practitioners to adopt methods. Therefore, we suggest studies in industry (e.g., surveys) to collect data on how practitioners handle variability of quality attributes in service-based systems. For example, results of our study help formulate hypotheses and questions for such surveys. Based on needs in practice, new approaches can be proposed.