FUPA: Future Utilization Prediction Algorithm based Load Balancing Scheme for Optimal VM Migration in Cloud Computing

Abstract

As a competitive model, Cloud Computing (CloudC) is fast growing model that allows users to access on-demand systems and web services. In CloudC, Load-Balancing (LB) is one of the most vital problems to allocate the workload uniformly among each node. Typically, LB defines the task of allocation and re-allocation of the workload among all the available resources to reduce the resource usages. For this task, different algorithms have been proposed. In recent year, an Osmotic Hybrid artificial Bee and Ant Colony (OHBAC) optimization algorithm has been proposed for achieving LB in the dynamic CloudC. In this algorithm, automated Virtual Machines (VMs) that are migrated via cloud systems were enabled by the osmotic behavior. But, this algorithm was only focused on minimizing the number of active Physical Machines (PMs) based on their current resource requirements and neglecting the potential resource requirements. This creates the unessential VM migrations and increases the Service Level Agreement (SLA) violations in data centers. Therefore in this article, OH-BAC algorithm with the Future Utilization Prediction (FUP) is proposed to reduce the amount of VM migrations and enhance the LB. In this newly proposed OH-BACFUP algorithm, both the current and future utilization of resources are considered to migrate the VMs into the least amount of active PMs. The future resource utilization is estimated by two different prediction models such as linear regression and optimal piecewise linear regression algorithms. These regression models can estimate the potential resource usages of VMs and PMs. Then, the predicted value can be used in the fitness function of OH-BAC to choose the best VM for migration to the most suitable PM.