Performance study of resource aware energy efficient VM placement algorithm

Abstract

Cloud computing is a computing paradigm that offers computing and infrastructure services over the internet. Cloud computing has changed the way resource management is done in data centers. Virtualisation technology, the backbone of cloud data centers, enables multiple virtual machines(VM) to be hosted on a single physical machine(PM). These virtual machines are easy to create, manage and remove from the physical machine in a matter of few seconds. Given the flexibility of working with VMs, the major problem is to identify the appropriate PM to host the VM. The process of mapping a VM request to a PM called VM placement problem is addressed by many researchers in different environments and metrics. The major bottleneck in adopting cloud computing is the cost incurred in terms of energy consumed by its data center. This paper aims to address the VM placement problem with an emphasis on minimizing energy consumption. It takes into consideration the resource characteristics such as CPU core and memory of the physical machine during VM placement. The performance of the algorithm is studied using the cloudsim toolkit. The performance metrics like the total energy consumption and the average resource usage of the VM requests with heterogeneous workloads, have been taken for the study. The result is compared with the standard First Fit algorithm and is presented.