Host load Prediction-based GMDH-EA and MMTP for Virtual Machines Load Balancing in Cloud Environment

Virtual machines (VMs) dynamic consolidation is effective to improve the utilization of resources and energy efficiency in cloud environment. However, the obligation of providing high quality of service to customers leads to the necessity in dealing with the energy performance trade-off, as aggressive consolidation may lead to performance degradation. Current solutions to the problem of host load detection are generally heuristic based. We propose a novel load balancing approach that combines the Group Method of Data Handling (GMDH) based on Evolutionary algorithm (EA) for host load prediction and the Minimum Migration Time policy (MMTP) for VMs migration. The GMDH-EA algorithm could predict the actual host load in each consecutive future time interval. We evaluate our method using the host load traces in the Google data centers with thousands of machines. The proposed algorithms significantly reduce energy consumption, while ensuring a high level of adherence to the Service Level Agreements (SLAs).