A Novel Approach for Dynamic Load Balancing with Effective Bin Packing and VM Reconfiguration in Cloud

Abstract

Background/Objectives: In cloud computing, the existing job scheduling algorithms were focused either on efficient job scheduling or optimal load balancing among the virtual machines. Methods/Statistical analysis: This paper introduces a novel approach called Dynamic Load Balancing with effective Bin Packing and VM Reconfiguration (DLBPR) in the cloud. In the proposed work, the jobs are initially classified using the deadline based job scheduler and stored in a different job queue based on the expected processing speed of the job. After classification, the jobs in the various queues are prioritized using their attributes. Findings: The proposed approach dynamically splits and coalesces the VMs based on the required processing speed of the job. The VMs in the data center are dynamically clustered based on their processing speed with the support of VM live migration, and the jobs are processed using the VMs in the cluster. Applications/Improvements: The proposed work is experimented in a cloudsim that minimizes the physical machine nearly 22% compared to other existing algorithms.