Agent based Scheduling Policies for Datacenter Environment

Abstract

Cloud computing is a modern technology  that allows the usage of IT infrastructures and services that do not seem to be installed on an area computer or server. The usage of clouds has become large in recent times. With this extensive usage, there are some issues arising because of inefficient job scheduling and resource allocation. These issues might cause dissatisfaction of users because of large waiting time to access resources and decline in cloud service providers' benefits. This might lead to high energy consumption, so as to eradicate this issue, an efficient job scheduling algorithm is proposed to cut back the delay time and energy consumption that tries to unravel the complexity involved during thescheduling phase. Main principle of this work is to make agents to act as a scheduler that dynamically assign VM resources for a user job within deadline. Experimental evaluation shows efficient execution of scheduling requests using agent in parallel manner than that of using traditional scheduling algorithm without agents.