Design And Implementation Of Distributed Architecture For E-commerce Platform

In 2018,the number of Internet users in China has reached nearly 800 million.The rapid development of the Internet coupled with the continued popularity of the mobile Internet has laid a solid foundation for the rapid development of e-commerce.Some e-commerce companies that are relatively well-known in the country are constantly adjusting their own business and technology deployment.Visible in the future of e-commerce competition will also become more and more intense,and e-commerce platform architecture design and research and development is even more a major challenge for corporate development.After analyzing and studying the technological changes in the development of e-commerce platform,this paper has reconstructed and implemented the existing architecture based on Tao Tao e-commerce platform.In the original architecture of the system,the invocation between subsystems adopts the older web service technology,which results in a high degree of coupling between the systems,resulting in poor reusability of the subsystems and high invocation cost between systems;After the system adds or modifies products,the search system's index database cannot be synchronized in real time.Based on the problems in the system,this paper reconstructs the original architecture of the system.The application that is used multiple times in the system is used as the service layer,the application program is used as the presentation layer,and the distributed service solution is introduced to implement the communication between the presentation layer and the service layer.The coupling between the systems is reduced,the reusability of the subsystems is improved,the inter-system call becomes simple,and the maintenance is easy.For the problem that new products cannot be synchronized in real time in the background,we propose to use the ActiveMQ mechanism to achieve synchronization between the index database and the database,so that the search system and the background management system are asynchronous operations,and the performance of the entire system will not be affected.,thus solving the problems of the original architecture.Based on the SOA architecture,this paper reconstructs the original system and adopts multiple clusters to achieve high availability.The foreground system uses the load balancing technology Nginx plus keepalived to achieve high availability and improve system concurrency.The background system adds the cache layer Redis between the user layer and the database layer,and adopts the Redis cluster method to reduce the database access pressure and improve the system performance.Using Redis solves the problem of distributed system session sharing.Maven manages the project and builds a picture server to complete the storage of large files.Full-text search uses SolrCloud clustering,and cluster management uses Zookeeper.