Design And Implementation Of Edge Computing Cloud Platform Based On Cloud-Edge Fusion

Recently,"cloud computing," "Internet of things," and other relevant technology have been widely used.But,with the rapid development of the 5G technology,its high speed,large capacity,and low delay make the terminal device produce the explosive growth of data that has brought our computing facilities safety,real-time and network dependence,and other challenges.In this context,the development of edge computing platforms has broad prospects.However,there are still some disputes about the development trend and specific implementation of edge computing platforms,such as where edge computing should sink,function positioning of cloud and edge terminal,and how to ensure the reliability of edge terminal.In addition,many existing platforms are too complex and have a long development cycle,which is not convenient for users to learn.Therefore,an edge computing platform that can effectively solve the above problems is needed.Compared with the existing edge computing platform,this thesis repositions the cloud and edge side functions in edge computing to make the function positioning of the cloud and edge clear.The server is used as the edge end to solve the insufficient computing capacity of the edge end.Docker is used for service deployment to make the platform more lightweight and ensure the reliability of the edge computing platform through intelligent management and monitoring of edge containers.Based on the multi-source synchronization technology and batch processing technology,the bidirectional automatic data synchronization between cloud and multiple edges ends is realized,which speeds up the synchronization speed and ensures the consistency of cloud-edge data.These innovative improvements have effectively solved the above problems.This thesis mainly discusses the design and implementation of the edge computing platform for cloud-edge fusion.The main goal is to unload some cloud functions to the edge processing.The platform provides three functional modules:request dynamic forwarding,intelligent container management,and monitoring,and bidirectional automatic synchronization of cloud-edge data.Through the request dynamic forwarding module,the request from the front end is forwarded to the corresponding cloud or edge end for processing according to forwarding rules,and the dynamic configuration of forwarding rules is realized.By container intelligent management and monitoring module,containers can be deployed to the cloud platform management or other edge sides,and provide the intelligent monitoring.At the edge of the container when there is a failure in the edge of the end,the cloud initiative running fault container corresponding image edge and handle originally sent to the request,the platform has more powerful ability of disaster.The platform periodically synchronizes the data from the edge to the cloud through the bidirectional automatic synchronization module of cloud-edge data.It automatically synchronizes the data generated from the cloud to the edge when the edge container recovers and runs,thus realizing the consistency of the corresponding data between the cloud and the edge.The primary technology adopted in this thesis is Java,the server-side mainly uses Spring Boot,Spring Cloud,and other relevant frameworks,and the relevant services are deployed through Docker,which is more lightweight.Swarm and Docker Java are used for container management,Mysql is used for database,and Beetlsql and Spring Data Jpa are used for persistence layer framework.After the preliminary test,the platform can effectively reduce the pressure on the cloud and reduce the response time of the front-end request.At present,the platform is in use,and all functions of the platform run smoothly.The use of the platform also effectively reduces the number of abnormal cloud server services and the response time of front-end requests.