A persuasive resource-aware allocation scheduler for enhancing task scheduling

Abstract

The deployment of Map Reduce has been built to grant enhancements to total system objectives such as job throughput. Hence, the support for user-specific objectives and resource allocation management has been least regarded and addressed. Schedulers enable users to assign jobs to queues that fulfil shared of specific resource. Existing work mainly focus on scheduling glitch occurring on the master’s side where the scheduler on the master node tries to allocate same work across all the worker nodes. The proposed scheduler focus on enhancing resource allocation when various kinds of workloads execute on the clusters. In order to evaluate the performance on the proposed scheduler which enhances resource utilization, an accomplishing time goal with each job is created.