A virtual machine scheduling method and scheduling monitor

Abstract

An embodiment of the invention discloses a virtual machine scheduling method. The method includes when a circumstance that central processing unit (CPU) occupancy rate of a first physical server is larger than or equal to a preset first threshold value is monitored, acquiring a first running log of each virtual machine in the first physical server, wherein each first running log includes a corresponding relationship between first resource using size of the corresponding virtual machine and time; calculating first resource using similarity among the virtual machines according to the corresponding relationship between the first resource using size of each virtual machine and the time, and acquiring a to-be-controlled virtual machine corresponding to the highest first resource using similarity; moving the to-be-controlled virtual machine out of the first physical server to lower the CPU occupancy rate of the first physical server. The embodiment of the invention further discloses a scheduling monitor. By the virtual machine scheduling method and the scheduling monitor, resource utilization rate of a physical server can be increased, and energy consumption can be lowered.

Description

technical field [0001] The present invention relates to the field of communication technology, in particular to a virtual machine scheduling method and a scheduling monitor. Background technique [0002] With the development of virtualization technology, virtualization technology has been widely used in data centers. Through virtualization technology, data centers can virtualize multiple isolated virtual machines on a physical server, and each virtual machine can run concurrently. Independent operating systems, these operating systems do not exclusively share the underlying physical computing resources, but run in a virtual machine that is exactly the same as the real computing environment, where exactly the same as the real computing environment refers to having the same CPU frequency, memory and HDD capacity, etc. Through the application of virtualization technology, the data center can cope with the increasing demand for data processing, can effectively improve the utili...

AI Technical Summary

Problems solved by technology

However, there is scheduling jitter in such scheduling, that is, when the scheduling monitor combines physical resource utilization information to analyze that the CPU usage of a physical server is very low at a certain point in time, the scheduling monitor migrates a virtual machine to the physical server to run. However, the low CPU occupancy rate of a physical server at a certain point in time can be a transient state, and the occupancy rate may become high in the later period, which may trigger rescheduling, resulting in low work efficiency of the physical server and energy consumption overhead

Images