Highly-Available Web Service Community

Functionally similar Web services are grouped as a community to facilitate and speed up the process of Web services discovery. This paper presents a solution to keep the Web service community highly-available to the user or application. The term highly-available refers that the Web service community can continue providing services even when master Web service (i.e., coordinator) fails operationally. Our solution customizes a distributed election algorithm called fast bully algorithm to identify a temporary master Web service when there is any operational failure in existing master Web service of community. The identified temporary master Web service handles the service provision and management responsibilities of Web service community. Permanent master Web service takes back the mastering responsibilities from temporary master Web service when it resumes. For this, we introduce some additional functions to the existing architectures of master and slave Web services to run the customized fast bully algorithm. Finally, a weather community is developed as a prototype example to illustrate our ideas practically.