Research On Micro-module Elastic Scheduling Technology Based On Edge Computin

With the further integration of the Internet and the real industry in recent years,a large number of edge terminal devices have been connected to the edge of the Internet.The edge scenario has high requirements on service delay and data security.In order to provide better services for edge scenarios,edge computing is derived.According to the different nature of services,edge cluster and edge node are used for deployment.However,all kinds of resources in edge scenarios are often limited.Therefore,in order to improve the utilization rate of edge resources and the balanced allocation of resources,this paper proposes an elastic scaling strategy for edge clusters and a micro-module resource scheduling strategy based on edge computing.Firstly,a micro-module elastic scaling strategy for edge cluster deployment is proposed.Based on the analysis of the problems existing in the traditional container cloud micro-module elastic scaling strategy in k3 s edge cluster scenario,an edge-oriented micro-module elastic scaling strategy combining active and passive strategies is proposed.At the same time,through the delivery of the cloud model,the elastic scaling strategy of edge cluster micro-modules can improve the resource utilization under the cooperation between edge clouds.Secondly,a micro-module resource scheduling strategy based on edge computing is proposed for the deployment scheme of edge nodes.By improving the default scheduling policy for the edge computing scenario,the delay filtering policy is added in the pre-selection phase,and the scoring policy based on resource dynamic load balancing and multi-resource requirements of edge task micro-modules are added in the optimization phase.Considering the resource dynamic load balancing requirements of the entire edge computing system and the specific resource usage requirements of a single edge micro-module task,the Scheduling Framework extension mode of the default scheduler is used to implement this method.Finally,based on the above research content,the experimental environment deployment is carried out.The experimental environment deployment of edge computing is carried out based on Openstack cloud platform of the central cloud and edge industrial computer,etc.,and the control group is set up to compare with the default strategy and the strategy proposed by others.After experimental testing,The micro-module elastic scaling strategy oriented to edge cluster can improve the system resource rate and reduce the response time.The micro-module resource scheduling strategy based on edge computing can realize dynamic load balancing on the premise that edge services meet the delay.