Load Balancing Scheme for optimization of Virtual Machine Migration using swarm in Cloud Environment

Abstract

In today's world, Cloud Computing is one of fastest growing model that gives access to the users for on demand service of resources and platforms. In Cloud Computing, Load Balancing is considered as one of the vigorous problems to distribute the work to various Virtual Machines (VM) uniformly among each node on the internet. Load balancing is defined as a job allocation and de-allocation of resources among various allocated resources for the reduction of resource usages. For enhancing load balancing, different authors have proposed various algorithms. In past few years, Osmotic Hybrid Artificial Bee and Ant Colony (OH-BAC) method has suggested aimed at optimizing load balancing in self-motivated cloud environment. In the proposed algorithm, VM that need to be migrated are permitted by osmotic behavior. However the drawback of above mentioned algorithm was that it only focuses on reducing the total number of Physical Machines (PM) based on existing requirements of the resources. Due to this reason, there are nonessential VM migrations and it will also affect the Service Level Agreement (SLA) violation. So, a new algorithm, OH-BAC with another algorithm Future Utilization Prediction, also known as FUP has been proposes to reduce VM migrations and focuses on enhancing Load Balancing onto the cloud. In the proposed algorithm OH-BAC-FUP, together existing and future utilization of resources must be measured as one the host for the migration of VMs from the existing number of lively PMs. The approaching source operation is projected by two distinct forecast models like linear regression and optimal linear regression algorithm. Hence, above stated regression models must be estimating the possible usage of resources of VM and PM. Then, the estimated value can be used in the fitness function of OH-BAC and from there it chooses the best VM that need to be migrated to suitable PM.