Abstract
A service oriented architecture platform consists of distributed web services (WSes). A system needs dynamic composition of WSes. Reliability of a composition process requires detection of faults and tolerance to the faults. A composition process can end abruptly due to occurrence of faults. Present paper proposes a quality of service (QoS) based fault-detection and fault-tolerance approach using the dynamic orchestration. The approach also considers the QoS of WSes and user preferences at runtime. The approach uses reliability in two-phases. Firstly, a trust based web-service filtering mechanism is used. This achieves reliability at the component level before a fault occurs. Next, whenever a fault detects in a process, a decision for dynamic recovery is taken. That is based on the optimum QoS ranking of the WSes. The steps in the proposed approach provide the reliability at the component as well as composition level. They work for all the orchestration based composition models. An implementation followed by experimental-study shows that the proposed approach produces timely as well as optimal results in the presence of the faults.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call