Workload Stability-Aware Virtual Machine Consolidation Using Adaptive Harmony Search in Cloud Datacenters

Owing to the increasing complexity of managing IT infrastructure caused by rapid technological advancements, organizations are transforming their datacenter management environments from on-premises to the cloud. Datacenters operating in the cloud environment have large amounts of IT infrastructure, such as servers, storage devices, and network equipment, and are operational on all days of the year, thus causing power overconsumption problems. However, efforts to reduce power consumption are not the first priority as datacenters seek stable operation to avoid violating their service level agreements. Therefore, a research model that reduces power consumption of the datacenter while enabling stable operation by utilizing virtual machine (VM) consolidation is proposed here. To obtain the optimal solution for the proposed VM consolidation model, an adaptive harmony search methodology is developed, which expends less effort to set the parameters of the model compared to existing harmony search methods. Comparative experiments were conducted to validate the accuracy and performance of the proposed model. As a result, Original harmony search (HS) showed better performance than the existing heuristic algorithm, and novel self-adaptive (NS)-HS showed the best result among Adaptive HS. In addition, when considering workload stability, it showed better results in terms of datacenters (DC) stability than otherwise.