Virtual machine scheduling management method and scheduler

Abstract

The invention provides a virtual machine scheduling management method and a scheduler. The method comprises the following steps of acquiring, by a scheduler, service characteristic information and/or state characteristic information of a virtual machine in a cluster, and managing the virtual machine according to the service characteristic information and/or the state characteristic information. In the above technical solution, the virtual machine in the cluster is scheduled automatically by the scheduler, so that compared with manual operation, the method makes the scheduler more intelligent and flexible; in addition, when managing the virtual machine, the scheduler dynamically manages the quantity of the virtual machines in the cluster according to load information and/or state characteristic information of the virtual machines in the cluster, thereby improving the use ratio of the system resource. Further, in the above technical solution, the scheduler can schedule the resource of virtual machines in the cluster in a unified manner.

Description

technical field [0001] The invention belongs to the field of virtual machine management, and in particular relates to a virtual machine scheduling management method and a scheduler. Background technique [0002] The cloud computing management platform project (OpenStack) can effectively manage computing, storage and network resources, greatly improving the utilization rate of hardware resources, and greatly improving the maintainability and scalability of business systems; the cloud computing management platform project (OpenStack ) provides advanced functions of Load Balance as a Service (LBaaS, Load Balance as a service), which improves system performance and reliability by distributing load among multiple virtual machines. [0003] However, when load balancing as a service (LBaaS) implements load balancing, the scheduling and management methods for virtual machines in the cluster mainly include the following points: [0004] 1. When a virtual machine in the cluster needs...

AI Technical Summary

Problems solved by technology

[0004] 1. When a virtual machine in the cluster needs to be added, it needs to be manually operated, which is not smart enough;

[0005] 2. When scheduling the virtual machines in the cluster through the scheduling policy pool, it does not take into account the dynamic management of the number of virtual machines in the cluster according to the load information, resulting in low utilization of system resources;

[0008] 1. Some aspects of virtual machine management (for example: adding a virtual machine) need to be manually operated, which is not smart enough;

[0009] 2. When scheduling virtual machines, dynamic management of the number of virtual machines in the cluster is not considered based on load information, resulting in low utilization of system resources;

[0010] 3. Scheduling policy pools and health check monitors are decision makers for managing virtual machines. There are too many decision makers and there is no unified cluster scheduling center, which is not conducive to unified scheduling of resources

Images