A Novel Reputation Management System for volunteer clouds

Volunteer computing systems provide many thousands of Teraflop´s of processor power for a wide range of scientific applications. There is a huge potential to use the contributed processor power for cloud computing applications. Cloud computing architectures have a robust design and are resilient to failures, whereas volunteer computing systems are highly distributed and prone to volunteer node failures. In order to deploy cloud services with high reliability in such volunteer clouds, we need a robust reputation management system to help estimate reputation of volunteer nodes for deciding on service deployment. This paper presents a Novel Reputation Management System for Volunteer Cloud. We have identified three basic trust parameters, namely, Performance in terms of time taken to do a job, Crashes of a node, and Correctness of returned results. We then define an overall reputation metric that combines the parameters to represent the overall reputation of the volunteer node. The paper also includes a discussion on possible strategies to implement the system in a distributed volunteer cloud.