Cluster distributed resource scheduling method, device and equipment and storage medium

Abstract

The invention discloses a cluster distributed resource scheduling method. The method comprises the steps of respectively acquiring the CPU utilization rate of each host and the CPU utilization rate ofeach virtual machine; quantizing the CPU utilization rate of each host and the CPU utilization rate of each virtual machine to obtain a quantized pressure value of each host and a quantized pressurevalue of each virtual machine; acquiring each historical host pressure value and each historical virtual machine pressure value; calculating a target CPU pressure value corresponding to each host according to the host quantized pressure value and the historical host pressure value corresponding to each host, and the virtual machine quantized pressure value and the historical virtual machine pressure value corresponding to each virtual machine in each host; and performing thermal migration operation on each virtual machine in each host according to each target CPU pressure value. According to the method, the pressure of network resources is reduced, and the probability of memory data loss caused by frequent thermal migration of the virtual machine is reduced. The invention further disclosesa cluster distributed resource scheduling device and equipment and a storage medium, which have corresponding technical effects.

Description

technical field [0001] The present invention relates to the technical field of distributed storage, in particular to a cluster distributed resource scheduling method, device, equipment and computer-readable storage medium. Background technique [0002] Cluster Distributed Resource Scheduling (DRS) is an important function of the virtualization platform. It can evaluate the resource pressure between hosts in real time, and schedule the virtual machines in the cluster according to various indicators, so as to achieve each cluster in the cluster. The host resources can be utilized in a balanced manner, avoiding the risk that the virtual machine cannot run normally or even go down due to the resource shortage of a certain host. [0003] The existing cluster distributed resource scheduling method is to evaluate the CPU resource utilization of all hosts in the cluster in real time, and then transfer the virtual machines in the hosts with high CPU resource utilization to the curren...

AI Technical Summary

Problems solved by technology

The advantage of this method is that virtual machines with high CPU load can be migrated to hosts with idle CPU resources in a timely manner according to the real-time CPU utilization. A large amount of CPU resources will be occupied during the execution process. In this case, the CPU utilization of the virtual machine will fluctuate periodically, causing the distributed resource scheduler to continuously perform hot migration of the virtual machine between hosts in the cluster. Such frequent live migration of virtual machines has many hidden dangers in data security, and it also puts a lot of pressure on network resources.

Images