Resource Matchmaking in Grid - Semantically

A fundamental task on the grid is to decide, which job is to run on which computing resources based on job or application requirements. The dynamic nature of the grid makes it difficult for the user to compile the complete list of available resources. Further matching the individual resources with requirements is error-prone task. In this paper we explain our work on developing such component called semantic resource broker. The semantic resource broker basically does resource allocation based on a policy. In our work, we consider resources with different capabilities. A resource is described in terms of parameters. The problem is to define resource allocation policy that allocates resources to jobs based on semantic matchmaking and that which decreases turnaround time. The main contribution of this paper is the development of semantic matchmaking policy that matches the request with the available resources using semantic technology and reduces the average turnaround time of jobs based on user preferences.