Virtual machine power consumption measuring method based on performance counter in cloud computation environment

Abstract

The invention discloses a virtual machine power consumption measuring method aiming at a cloud computation system. The measuring method comprises the steps of firstly selecting out a performance counter set most representative in power consumption, then constructing a mapping relationship between a performance counter and physical device power consumption and a model by use of a multi-order successive approximation method, and thus obtaining a total power consumption measuring model of a virtual machine. The virtual machine power consumption measuring method has the advantages that the accuracy of the power consumption model of the virtual machine is increased remarkably by use of the method compared with that by use of the existing method, the sampling time accuracy in power consumption measurement is improved from a millisecond level to a microsecond level and therefore an effective support is provided for realizing fine-grained power consumption control and management of a cloud system.

Description

technical field [0001] The invention relates to the field of cloud computing, in particular to a method for measuring power consumption of a virtual machine and its implementation technology. Background technique [0002] Cloud computing system has gradually become the most promising distributed computing platform. In order to ensure the service quality of cloud systems, servers in data centers often need to provide users with access to resources 24 hours a day, and this "over-supply" resource sharing method leads to huge energy consumption. Therefore, energy consumption control and optimization management for cloud systems has become a research topic with great application value and commercial value. [0003] In a cloud computing environment, high-performance data centers often provide services for upper-level users through resource virtualization technology, that is, the basic unit of resource allocation and management is an independent "virtual machine". Therefore, the ...

AI Technical Summary

Problems solved by technology

The main disadvantages of this technology are: low accuracy, long modeling cycle of power consumption model, and obtained power consumption model can only be applied to specific physical devices

The main disadvantages of this technology are: the overhead of establishing and maintaining the "feature-power consumption" database is very high, the efficiency of the pattern matching algorithm will restrict the real-time performance of power consumption measurement, and only approximate power consumption can not be obtained. Accurate power consumption value

Its main disadvantage is: the construction time of the power consumption model is very long, and it cannot meet the requirements of real-time power consumption measurement

Images