Anomaly detection method of internal virtual machine of cloud system

Abstract

The invention relates to an anomaly detection method of an internal virtual machine of a cloud system, which comprises the following steps: state information of a normal virtual machine of the cloud system is collected to train Hidden Semi-Markov Model (HsMM), and a corresponding algorithm is designed for detecting and computing probabilities and mahalanobis distances of resource dynamic change behaviors when various virtual machines in the cloud system are on line; if the mahalanobis distance in the online detection result of some virtual machine is more than a preset threshold value, the activities of the virtual machine are abnormal, so an anomaly detection and processing system in the cloud system is started for anomaly detection and processing of the virtual machine; if the anomaly rate of some virtual machine is detected to be less than the maximum threshold value of the anomaly detection and processing, after the anomaly is eliminated, a warning prompt is sent to a cloud tenant; otherwise, an alarm is given for the cloud tenant of the virtual machine, and the virtual machine is closed. According to the anomaly detection method, abnormal behaviors of the internal virtual machine of the cloud system can be detected in real time, the occupied system resources are less, and the high availability and high safety of the internal virtual machine of the cloud system can be fully ensured.

Description

technical field [0001] The invention relates to the field of network technologies, and in particular to an abnormality detection method for a virtual machine in a cloud system. Background technique [0002] More and more companies and enterprises are reducing their costs by migrating parts of their information technology infrastructure to cloud service providers, such as the widespread use of data centers with distributed storage infrastructure and other types of cloud computing systems. Cloud service providers use commercial virtualization software such as Vmware and vSphere to build various types of virtual infrastructure, including private cloud and public cloud systems, whose data may be distributed among hundreds of interconnected computers, storage devices and on other physical machines. [0003] In a public cloud system or a private cloud system, the enterprise rents the computing resources and storage resources of the cloud service provider, that is, the cloud tenan...

AI Technical Summary

Problems solved by technology

Detecting the existence of abnormal virtual machines poses a challenge to cloud security. At present, there are relatively few anomaly detection and processing methods for virtual machines in cloud systems, and the existing defense technologies do not take into account the dynamic changes in the activities of virtual machines in cloud systems. Therefore there are certain limitations

Images