Comparison between Light-Weight and Heavy-Weight Monitoring in a Web Services Fuzzy Architecture

Web services are a reality for many businesses nowadays. The rules of these services are described on e-contracts (ECs). Therefore, monitoring is essential to ensure that the quality of service will be higher than agreed in the EC. This paper proposes an architecture for business process execution, where the monitor uses a fuzzy approach to predict an EC failure, and take actions to avoid it. With this prediction, the architecture changes service priority by running services with higher possibility (higher priority) of failure first. Nevertheless, if a failure happens, the architecture has a recovery module to recovery the service. Using the architecture, it is possible to observe an increase in the EC accomplishment (+40.41%), and a decrease in the average response time of EC (-42.64%). This paper compare two types of monitoring: light-weight monitoring (LWM) and heavy-weight monitoring (HWM). The results show that HWM is better than LWM in terms of performance. There was an improvement of 11.88% in the EC accomplishment. The problem using the HWM is the reliability. It is not reliable. If the monitor fails, no monitoring is processed by the architecture. The results show that the architecture is promising. (C) 2015 The Authors. Published by Elsevier B.V.