基于工作流的无服务边缘计算框架设计与实现  

Design and Implementation of an Improved Serverless Edge Computing Framework Based on Workflow

作　　者：高明[1] 许霖卓 GAO Ming;XU Lin-zhuo(School of Information and Electronic Engineering,Zhejiang Gongshang University,Zhejiang Hangzhou 310018,China)

机构地区：[1]浙江工商大学信息与电子工程学院,浙江杭州310018

出　　处：《计算机仿真》2024年第4期461-465,共5页Computer Simulation

基　　金：国家自然科学基金资助项目(61871468);国家自然科学基金国际合作与交流项目(62111540270);浙江省基础公益研究计划(LGG20F010015)。

摘　　要：边缘计算具有高性能、体积小、计算力超强等优点。但云设计的边缘计算函数链存在调用时延较长的问题,导致大规模部署和存储能力以及云端和设备端协同困难、无缝对接效果较差。为解决上述问题,基于开源无服务平台OpenFaaS设计了一个无服务边缘计算框架(SECP),为边缘侧提供基于无状态函数的计算服务。对计算服务调用过程采用建模分析,采用了工作流模型对为云设计的无服务计算在边缘侧存在的函数链式调用时延放大的问题改进。性能测试结果表明,SECP在边缘计算场景下有良好的性能表现,降低了边缘计算平台中服务请求的响应时间,解决了无服务计算函数功能单一问题,提高了边缘侧资源的利用率,并改善了边缘侧计算服务部署、维护繁杂的问题。Edge computing has many advantages,such as high performance,small volume and strong computing power.However,the edge computing function chain designed by the cloud has a long call delay,which may lead to difficult collaboration between the cloud and the device,and an unsatisfactory seamless docking effect.Therefore,this paper designed a serverless edge computing framework(SECP)based on OpenFaas,which provides computing services based on stateless functions for the edge side.Firstly,the invocation procedure of the computing service was modeled and analyzed,and then the workflow model was used to improve the problem of function chain invocation delay amplification existing on the edge side of the serverless computing designed for the cloud.The test result shows that SECP has good performance in edge computation,reduces the response time of service requests in the edge computing platform,solves the problem of single function of serverless computing,improves the utilization rate of edge resources,and improves the problem of complicated deployment and maintenance of edge computing services.

关 键 词：无服务计算 边缘计算 工作流 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]