A decentralized utility-based grid scheduling algorithm

Grid systems have gained tremendous importance in past years since application requirements increased drastically. The heterogeneity and geographic dispersion of grid resources and applications place some difficult challenges such as job scheduling. A scheduling algorithm tries to find a resource for a job that fulfills the job’s requirements while optimiz-ing a given objective function. Utility is a measure of a user’s satisfaction that can be seen as an objective func-tion that a scheduler tries to maximize. Many utility func-tions have been proposed as an objective for scheduling al-gorithms. However, the proposed algorithms do not con-sider partial requirement satisfaction by awarding an utility based on the total fulfillment of the requirement. Most of them follow centralized or hierarchical approaches, suffer-ing from scalability and fault tolerance problems. Our so-lution proposes a decentralized scheduling architecture with utility based scheduling algorithm that considers partial re-quirements satisfaction to overcome the shortcomings of ac-tual solutions. Performance results show that user utility, submission and execution times are improved and a slightly more balanced system is achieved