Research And Implementation Of Online Mall Based On Dubbo Distributed Architecture

In recent years,e-commerce has developed rapidly at home and abroad,and the number of online shopping has increased year by year.The traditional framework structure can not meet the high concurrent demand,and the emerging Hadoop ecosystem can not completely replace the traditional project development.In order to solve the above problems,the distributed idea is introduced into the WEB development project to realize a distributed,highly concurrent,load balanced large and medium-sized e-commerce online mall system.The system has the characteristics of unification,standardization,modularization and flexible deployment.Based on the idea of service-oriented,this paper applies Dubbo Open Source Distributed Architecture,which is based on RPC communication technology,to coordinate service Zookeeper through distributed applications,and to complete the management and scheduling between service producers and service consumers.In the specific implementation framework,because the online mall system belongs to the Internet industry,it needs to reduce the correlation between functions.The system is developed with three framework models,Spring MVC,Spring and Mybatis,which are more lightweight.According to the two characteristics of IOC and AOP of Spring framework,the other two frameworks are integrated and managed by Spring container.In the aspect of data persistence,Mybatis framework's reverse engineering technology is used to access MySQL database.This framework can directly write SQL statements,making the development more simple and lightweight.In order to meet the actual needs,all the projects in the system are built based on Maven,and the key construction and dependency management of the project are carried out.An online shopping system supported by Dubbo architecture is implemented by applying advanced open source frameworks and technologies,such as Redis caching technology,Active MQ asynchronous message queuing technology and Solr full-text retrieval technology.This paper analyses and designs the system according to the technology and method mentioned above,and provides a good solution for the distributed implementation of the system.Decrease the coupling degree between systems as a whole,and improve the concurrency and scalability of the system.