Cloud server elastic expansion and contraction and performance optimization method based on model predictive control

Abstract

The invention relates to a cloud server elastic expansion and contraction and performance optimization method based on model prediction control. The method comprises control over server power consumption and control over virtual machine application level performance, a power controller in a cluster level power control loop adjusts the CPU frequency of each server through dynamic voltage and frequency adjustment (DVFS), and the total power consumption of all servers in a cluster is dynamically controlled. A performance controller in the performance control loop dynamically controls the application program performance of the virtual machine by adjusting CPU resources allocated to the performance controller. Because the total power consumption of the cluster needs to be kept lower than the capacity of the shared power supply, and the cluster-level power conversion between different servers can bring better system performance. By means of the cloud server elastic stretching and performanceoptimization method based on model prediction control, system power consumption can be reduced, it is guaranteed that the performance of the application program on the virtual machine meets certain requirements, and the service quality of the system is improved.

Description

Technical field [0001] The invention relates to a method for optimizing the power consumption and application-level performance of a virtualized server, in particular to a large-scale deployment of a virtualized computer system, such as a coordinated control method for the performance and power consumption of a server cluster in a data center. Background technique [0002] With the development of cloud computing, big data, machine learning and other technologies, users have increasingly greater demands for data storage, processing, and intelligent analysis. Through server virtualization technology, data centers can provide more flexible resources to carry more Application services and improve the manageability of services. However, as the scale of the server system of the data center increases, the energy consumption of the server increases, which not only increases the operating cost of the data center, but also affects the reliability of the server system and the quality of ser...

AI Technical Summary

Problems solved by technology

However, the existing research on the control management of power consumption and quality of service is not comprehensive, and some only consider the unilateral influence factors of two levels of power consumption application request performance, and cannot provide clear guarantees for both at the same time

On the other hand, although many current researches have proposed many control management scheduling strategies, such as effectively reducing server power consumption by switching hardware components to low-power states, most of these strategies are difficult to be directly applied to actual On the cluster-level server data center based on virtualization technology, and some strategies are limited by the commercial development of hardware manufacturers

Images