Computational Modelling of Collaborative Resources Sharing in Grid System

Abstract

In grid computing, Grid users who submit jobs or tasks and resources providers who provide resources have different motivations when they join the Grid system. However, due to autonomy both the Grid users' and resource providers' objectives often conflict. This paper proposes autonomous hybrid resource management algorithm for optimizing the resource utilization of resources providers using “what-yougive-is-what-you-get” Service Level Agreements resource allocation policy. Utility functions are used to achieve the objectives of Grid resource and application. The algorithm was formulated as joint optimization of utilities of Grid applications and Grid resources, which combines the resource contributed, incentive score, trustworthiness and reputation score to compute resource utilization. Simulations were conducted to study the performance of the algorithm using GridSim v5.0. The simulation results revealed that the algorithm yields significantly good result because no user can consume more than what it contributes under different scenarios; hence the problem of free riding has been addressed through this algorithm.

Keywords: Resource scheduling, Grid System, Computational modelling