Quality of Service Conflict during Web Service Monitoring: A Case Study

Abstract

Web services have become one of the most used technologies in service-oriented systems. Its popularity is due to its property to adapt to any context. As a consequence of the increasing number of Web services on the Internet and its important role in many applications today, Web service quality has become a crucial requirement and demanded by service consumers. Terms of quality levels are written between service providers and service consumers to ensure a degree of quality. The use of monitoring tools to control service quality levels is very important. Quality attributes suffer variations in their values during runtime, this is produced by many factors such as a memory leak, deadlock, race data, inconsistent data, etc. However, sometimes monitoring tools can impact negatively affecting the quality of service when they are not properly used and configured, producing possible conflicts between quality attributes. This paper aims to show the impact of monitoring tools over service quality, two of the most important quality attributes - performance and accuracy - were chosen to be monitored. A case study is conducted to present and evaluate the relationship between performance and accuracy over a Web service. As a result, conflict is found between performance and accuracy, where performance was the most affected, because it presented a degradation in its quality level during monitoring.