基于边缘计算的工业物联网容器管理引擎设计
2021-08-06郭家伊
郭家伊



摘 要: 针对工业物联网中云端压力大、工业协议标准繁多等问题,设计了一款基于边缘计算的工业物联网容器管理引擎。该引擎具有三个分布式端,分别部署了容器管理引擎KubeEdge、协议转化工具EMQ X Edge和可视化工具OCP、Kuboard,提供资源监控、镜像管理、持续集成、自动伸缩、协议转换等功能。测试显示,集群数据带宽提升339.19%,响应时间减少81.22%。把部分计算任务从云端卸载到边缘后,系统能源消耗减少30%-40%,成功解决带宽不够、云端压力大等问题。
关键词: 边缘计算; 工业物联网; 容器; KubeEdge
中图分类号:TP311 文献标识码:A 文章编号:1006-8228(2021)07-38-04
Design of the container management engine for industrial Internet of Things
based on edge computing
Guo Jiayi
(Hangzhou Dianzi University, Hangzhou, Zhejiang 310018, China)
Abstract: Aiming at the problems of high cloud pressure and various industrial protocol standards in industrial Internet of Things, a container management engine for industrial Internet of Things based on edge computing is designed. The engine has three distributed terminals, which respectively deploy the container management engine KubeEdge, the protocol conversion tool EMQ X Edge and the visualization tools OCP and Kuboard to provide resource monitoring, image management, continuous integration, automatic scaling, protocol conversion and other functions. The test results show that the cluster data bandwidth is increased by 339.19%, and the response time is reduced by 81.22%. After some computing tasks are unloaded from the cloud to the edge, the system energy consumption is reduced by 30% - 40%, which successfully solves the problems of insufficient bandwidth and high cloud pressure.
Key words: edge computing; industrial Internet of Things; container; KubeEdge
0 引言
萬物互联给人类带来了极大便利的同时,海量设备的通讯也受限于不同设备的物理链路、各种不同的协议,因此大多数数据不能互联互通[1]。同时由于目前主流的容器编排引擎,如Kubernetes、Mesosphere、Docker Swarm等[2],难以实现自动伸缩应用,缺乏对容器的统一管理[3],易造成带宽不够、实时性运维服务质量下降及企业成本增加等问题[4]。
在此情况下,传统云计算模式无法高效地支持基于万物互联的应用服务程序[5],而边缘计算模型可以更好地解决这些问题。目前,有研究者已开展相关研究,如潘建胜曾以Docker为基础,设计并实现了一个CDN边缘计算平台[6];王健通过轻量级的边缘节点构建轻量级边缘计算平台[7],实现了轻量级的虚拟化与网络交换;吴鸿飞通过模型和算法扩展了Kubernetes,开发了一个轻量级的边缘计算平台[8]。这些研究说明,边缘计算模型可以补充云计算在实时性、智能性和……
