Time efficient dynamic threshold-based load balancing technique for Cloud Computing

Cloud computing is a novel technology leads several new challenges to all organizations worldwide. Cloud computing supports virtual machines (VMs) to host multiple applications simultaneously. Balancing the large numbers of applications in the heterogeneous cloud environment becomes challenging as the hypervisor scheduling controls all VMs. When the scheduler allocates tasks to the overloaded VMs, the performance of the cloud system degrades. In this paper, we present a novel load balancing approach to organizing the virtualized resources of the data center efficiently. In our approach, the load to a VM scales up and down according to the resource capacity of the VM. The proposed scheme minimizes the makespan of the system, maximizes resource utilization and reduces the overall energy consumption. We have evaluated our approach in CloudSim simulation environment, and our devised approach has reduced the waiting time compared to existing approaches and optimized the makespan of the cloud data center.