Job Scheduling Algorithm based on Dynamic Management of Resources Provided by Grid Computing Systems

Abstract

The need of tasks scheduling has occurred to many real life situations. Within a grid computing system, the processing requests are randomly received from the system’s users. A good scheduling of these requests assumes the assignment of requests to available resources, so that all requests will be solved at once as possible. In this paper, an algorithm of tasks scheduling within a GRID system is proposed, by means of using the information come out of each GRID community resource. This algorithm meets a load balancing diagram, where the communication between nodes of the grid is accomplished through a mutual feedback of information. The nodes performing the information exchange are randomly chosen. The proposed algorithm is tested by a GSSIM simulator, where a comparison between this algorithm and other algorithms of tasks scheduling is carried out.