Improving performance of cloud datacenters using heuristic driven VM migration

Abstract

Cloud Computing has emerged as an alternate infrastructural, computational and service platform. It caters the users with least financial burden. Cloud uses virtualization for resource management through Virtual Machine (VM). An important aspect of cloud computing remains Quality of Service (QoS), often referred as Service Level Agreement (SLA). Several user applications simultaneously compete for resources in data centers. With limitations in available resources and compelling SLAs, VMs have to be migrated live during execution. Another concern in Clouds remains energy efficiency because of huge sized servers and communication requirements; most of the investment goes in power. In this paper, a novel SLA conscious variance based VM Selection scheme has been proposed. Simulation results revels that proposed scheme outperforms other approaches based on resource optimization for cloud computing environments. During simulation some real world workload traces from PlanetLab has been used.