Dynamic resource monitoring and scheduling method of cloud computing platform IaaS layer

Abstract

The invention relates to the technical field of cloud computing, and provides a dynamic resource monitoring and scheduling method of a cloud computing platform IaaS layer, for monitoring and scheduling resources of a data center composed of a control server and a plurality of computing servers. The method comprises: the control server collecting current state information of each computing server in the data center; traversalling and collecting use states of all working nodes in the data center; determining load states of the working nodes in the data center; performing virtual machine migration condition determination; selecting target virtual machines in the working nodes to be migrated; selecting target servers for the target virtual machines; and migrating the target virtual machines to determined target servers. According to the invention, a time-based SLA integration grade is utilized to measure the computing capacity of each server in the data center, the size of a memory and a bandwidth is taken as a consideration factor of resource scheduling, and the resource scheduling of the heterogeneous data center is finished. According to the memory occupation size of the virtual machines, a scope segmentation method is used so as to effectively measure the scheduling cost of each virtual machine.

Description

technical field [0001] The invention relates to the technical field of cloud computing, in particular to a dynamic resource monitoring and scheduling method of the IaaS layer of a cloud computing platform. Background technique [0002] With the rapid development of cloud computing technology, the various resources of the data center in the cloud environment are also increasing. How to reduce the energy consumption of the data center through resource scheduling and improve the utilization of system resources is a very important content. . At present, resource scheduling methods aimed at reducing data center energy consumption mainly include two types of methods: the first type mainly achieves energy saving by dynamically adjusting the voltage or frequency of the CPU; the second type is to shut down unnecessary servers resources to achieve energy saving. At present, the main strategy adopted by the scheduling method aimed at improving the utilization rate of system resources...

AI Technical Summary

Problems solved by technology

However, the reality is that the server configurations in the data center are often different, so the effect of this scheduling method is not ideal

This method only considers the computing power of the server, and does not fully consider the memory and network resources

At the same time, this scheduling method does not take into account the overall impact of scheduling costs and scheduling time on scheduling

Finally, this method does not take into account the impact of the server's instantaneous load change on resource scheduling, so it is prone to instantaneous load peaks or valleys to trigger unnecessary scheduling, resulting in waste of resources and affecting overall performance.

Images