Resource scheduling method and device

Abstract

The invention discloses a resource scheduling method and device. The method includes the steps that health status information of all hosts in a current environment is monitored and input in a pre-defined health status data model to obtain health status data of all the hosts; resource utilization rate data of all the hosts in the current environment is acquired; the hosts with the health status data and the resource utilization rate data being within a preset range are screened out to serve as standby hosts, and a scheduling preparation instruction is generated; the resource utilization rate data and the health status data serve as reference objects, target hosts are selected from the standby hosts, the target hosts are added into a target host list, and a scheduling instruction is generated in combination with resource scheduling information; the resource scheduling process is executed according to the scheduling instruction. From the top, the health status data and the resource utilization rate data of the hosts are comprehensively considered, the problems caused by the fact that only the parameter of the low resource utilization rate is considered can be effectively avoided, the reliability of resource scheduling is improved, and the risk of a crash is reduced.

Description

technical field [0001] The present invention relates to the technical field of cloud computing, in particular to a resource scheduling method and device. Background technique [0002] Currently, most server virtualization products on the market provide advanced virtualization functions such as virtual machine migration and dynamic resource scheduling. However, in the design of virtual machine migration and dynamic resource scheduling functions, usually only the utilization rate of the physical host CPU, memory, storage, network, etc. is used as the criteria for selecting the migration destination. The traditional method of selecting the destination host only by the utilization rate of the host's CPU, memory, storage, network, etc. is feasible in most cases, but in some abnormal cases, such as when a host fan is abnormal, we will When the business is migrated to this host, the application will face the risk of downtime. Even if the virtualization is protected by a high-avail...

AI Technical Summary

Problems solved by technology

The traditional method of selecting the destination host only by the utilization rate of the host's CPU, memory, storage, network, etc. is feasible in most cases, but in some abnormal cases, such as when a host fan is abnormal, we will When the business is migrated to this host, there will be a risk of downtime when applying it. Even if the virtualization is protected by a high-availability mechanism, there will still be a short-term business interruption

Images