Bayesian Networks-Based Selection Algorithm for Virtual Machine to Be Migrated
Chengyu Yan, Zhihua Li, Xinrong Yu, Ning Yu · 2016
In cloud data centers, virtual machine (VM) consolidation is one of the challenge topics. In which, the selection of VMs to be migrated is one of the key issues in the process of VM consolidation. In this paper, under consideration of the dynamical uncertain environment, a Bayesian networks-based estimation model was constructed. Because excessive VM migrations influence the Quality of Service (QoS) of data center, the model aims at estimating the migration probability of VMs and calculating the potential total number of migrations occurred in physical hosts. Based on the proposed model, a Bayesian networks-based selection algorithm (BN-SA) for VMs to be migrated was proposed. The BN-SA adaptively adjusts the overloaded threshold and selects VMs which have relatively short migration time and big impact on potential migrations of host in the phase of reallocating VMs. The experimental results show that BN-SA algorithm has a promising performance.