An Intelligent Anomaly Detection and Reasoning Scheme for VM Live Migration via Cloud Data Mining
Qiannan Zhang, Yafei Wu, Tian Huang, Yongxin Zhu · 2013
Cloud computing operators provide flexible, convenient, and affordable means to access public and private services. Virtual machine (VM) live migration, as an important feature of virtualization technique in cloud computing, ensures high efficiency and performance of computing infrastructure, while it stays transparent to clients. However, VM live migration is observed to cover anomalies due to their statistical similarity. To tackle the critical security issue, in this work, we propose an intelligent scheme to mine statistical data from cloud infrastructure to detect anomalies even if VMs are migrated to a new host with different infrastructure settings. In addition to detection of the existence of anomalies, our scheme is capable of identifying the possible sources of anomalies, which gives administrators clues to pinpoint and clear the anomalies.