Compensation Based Grid Scheduling

Abstract

In the grid-computing environment, resources are autonomous, distributed with varying pricing policies. The grid comprises of the contributors and the users who each play a role for certain compensational benefits. The proposed model is a tender contract type that envisages a real market operating on the supply and demand perspectives of a defined commodity. It allows a decentralized selection of contributors by the user for the execution of his job. The goal is to compensate the contributor for his resource and the user to pay less for his job. The contributor benefits by profits and fairness, which is evident as pricing of the resource based on its demand and efficiency. The user is compensated with the successful completion of jobs and fine amount borne by the contributor for not guaranteeing the defined deadline requirements. Simulation and results show that both the contributors and the users are compensated equally well