Towards Dependable Web Services in Collaborative Environments Based on Fuzzy Non-functional Dependencies

Web Services composition and collaboration has been an active area for researchers and organizations for its ease of inter-operation and application integration. Web Services composition allow creating complex business processes on which organizations rely on. Yet, the unreliable nature of the Internet and communication channels arise many challenges. Therefore, many studies are oriented towards solving issues like dependability and resilience of systems, some implementing fault tolerance and others for systems' reliability, issues that remain as an open challenge. For sake of mitigating the aforementioned one can implement the autonomic computing paradigm. However, attending to changing conditions and requirements, still a better diagnosis technique for observations made, and for learning, and to update functioning conditions, therefore guide the autonomic management of the system being manged. In this paper we adopt a diagnostic approach based on fuzzy logic, we define fuzzy nonfunctional requirements to make inferences, can take corresponding actions adapting to changes in the system under different workload conditions. We target improving the functionalities of Web Services platform integrating monitoring and fault diagnosis capabilities to them. We present experimental results to prove the effectives of our approach.