Web services give a new view of the web as the biggest, widely accepted and the most straightforward distributed software platform. Their composition into applications and business processes is still a complex, non-trivial task, requiring highly rational efforts not only from the software developers, but from the quality assurance specialists. The provision of web service compositions' quality brings a lot of challenges due to variability of difficulties at infrastructure, service and choreography levels and the need of different types of testing in unknown context and environment. A consolidated quality assurance methodology that advances the fundamental understanding of testing in terms of concepts, models, techniques, standards and automation is required. This methodology needs to enable effective exploration, comparison, evaluation and selection of testing approaches, platforms and tools. This article proposes such a methodology and reviews the current testing approaches for single and composite web services from an objective, holistic perspective. The methodology is presented as an end-to-end testing procedure, in which activities are facilitated by a set of testing approaches, techniques and best practices. A concrete solution is recommended for actual implementation of each activity either through selection among the most appropriate and effective existing approaches or development of new approaches, mainly in case of critical issues such as dependencies analysis, partner web services' isolation and injection of faults. A common framework that integrates different testing tools automates the methodology. Its applicability, completeness, level of automation, and level of novelty is evaluated through testing of real testing scenarios.
CITATION STYLE
Petrova-Antonova, D., Manova, D., & Ilieva, S. (2020). Testing web service compositions: Approaches, methodology and automation. Advances in Science, Technology and Engineering Systems, 5(1), 159–168. https://doi.org/10.25046/aj050121
Mendeley helps you to discover research relevant for your work.