AN IMPROVED FAULT TOLERANT SCHEDULING ALGORITHM FOR GRID ENVIRONMENT

Abstract

Grid computing is becoming popular due to the stupendous growth in the internet, bringing the resources together which makes the user to get fast solutions for the large scale problems. This enables sharing, selection and aggregation of a wide variety of geographically distributed resources. As size of the applications grow, more usage of resources for longer periods of time, may lead to increasing number of resource failures. When failures occur, the execution of the jobs that is assigned to the failed resources will be affected. So, fault tolerance is necessary in such cases. To overcome the failure, a scheduling algorithm is proposed that depends on a new factor called scheduling indicator in selecting the resources. This factor comprises of the response time and the fault rate of grid resources. The fault rate is based on the success and the failure of job execution. Whenever a grid scheduler has jobs to schedule on grid resources, it uses the scheduling indicator to generate the scheduling decisions. The resource with minimum scheduling indicator value receives the job.