SLAV Mitigation and Energy-Efficient VM Allocation Technique Using Improvised Grey Wolf Optimization Algorithm for Cloud Computing

A cloud computing environment is an attractive platform to facilitate large-scale sales processes for customers from various industries. However, extensive research has been carried out on the cloud data center to reduce power consumption due to operating costs and CO2 emissions. To provide a prominent solution for reducing power consumption in the cloud center, we proposed a novel VM allocation technique that overcomes the Service Level Agreement (SLA) violation and allocates the VM to the best host through Improvised Grey Wolf Optimization (IGWO) Algorithm. The proposed VM allocation technique takes into account host characteristics such as power consumption and CPU utilization to evaluate the best fit. After VM allocation, we evaluate the remaining capacity of CPU and RAM of the host, which enhances the resource utilization in the cloud center. The experiment was carried out on random datasets with different VMs. The performance of the proposed VM allocation technique is compared to the existing VM migration algorithm like Ant Colony Optimization (ACO) and PABFD (Power-Aware Best Fit Decreasing), the result shows that the proposed method provides a significant performance in terms of energy consumption, SLA violation, and several VM migrations. The proposed VM allocation method consumes low power and violates fewer SLAs, thus providing a green computing environment.

I. Introduction

Over the past few years, cloud computing has provided an efficient cloud computing platform for businesses, healthcare, and smart systems. Various other services are also available, including security for databases, network management, and software management. Customers can take advantage of a highly flexible and dynamic virtualization resource offered through cloud computing. There is wide research being conducted to determine how to properly implement task scheduling to maximize the customer's resources. There is still an NP-hard problem that cloud computing platforms must deal with, which calls for the use of new innovative algorithms. Datacenters are allocated to the cloud, which is utilized globally, where each data center has thousands of servers, and each server is divided into multiple virtual machines. Many enterprises continue to utilize the cloud during the pandemic period [1] and are ready to spend money on multi-cloud and hybrid clouds. Enterprises spend Thirty-six percent of annual spending on cloud computing, and Eighty-three percent spend more than $1.2 million per year on cloud computing. Ninety-two percent of enterprises consider utilizing multi-cloud strategies, while 82 percent consider hybrid cloud strategies. To provide prominent services to the users, the datacenters need enormous electric power, increasing a wide operating cost. Cloud centers consume approximately 1.3% of electric power worldwide which randomly a 9% increase in 2020.