Scheduling in Grid to Minimize the Imposed Overhead on the System and to Increase the Resource Utilization

Abstract

Grid computing is one of the area which is having outstanding advancement nowadays. Grid computing have emerged to harness distributed computing and network resources to provide powerful services. The non-deterministic characteristic of the resource availability in these new computing platforms raises an outstanding challenge. Since grid computing is rapidly growing in heterogeneous environment it is used for utilizing and sharing large scale resources to solve complex problems. One of the most challenging task in the area of grid is resource provisioning. As it is very well known that there are enormous amount of resources present in the grid environment, it is very hard to schedule them. There are two types users namely the local users and the external users. Resources has to be provided to both the users request. When the user submits the job, the request has to be satisfied by providing the resources. Here the resources are provided in the form of virtual machines (VM).The problem arises when there is insufficiency of resources to provide both the users. The problem becomes more hectic when the external requests have more QoS requirements. The local users can be provided with the resources by preempting the VMs from the external requests which causes imposed overhead on the system. Now the idea is how to decrease the preemption of VMs and how to utilize the resources efficiently. Here we propose a policy in Intergrid which reduces the no of VM preemption and which will also distribute the external requests accordingly so that preemption of valuable requests is reduces.