| The combination of self-service stores and fresh food e-commerce platforms has spawned a new retail concept of online ordering and offline self-service,which provides the possibility of contactless purchase of fresh goods in the context of the COVID-19.The contract fulfillment system of fresh commodities mainly uses the means of informatization of contract fulfillment process to deliver fresh commodities to consumers on time without omission.At present,it faces many problems and challenges,such as the high growth of new users and high concurrent orders of fresh e-commerce platforms during the epidemic period;The function of contactless self pick-up point stores to quickly distribute high-quality fresh goods nearby is not clear;There is a problem of repeated development,which requires a high compatibility platform to integrate the existing fresh e-commerce functions and abstract the micro services for commodity performance;To deal with the above problems,this paper designs and implements a fresh commodity performance system based on micro services.Based on the demand analysis of fresh commodity performance business,the system is grouped into four micro service modules: performance pipeline,performance preparation,performance production and performance configuration.Help users complete the whole process of fresh commodity performance including mobile phone order,store positioning,task distribution,commodity sorting,packaging and distribution.The system mainly supports and cooperates with the fresh e-commerce platform to complete the whole commodity performance process by providing callable interfaces to APP clients,PDA devices and WEB terminals.Therefore,during development,the back-end uses the Spring Cloud framework to implement the micro service architecture,and each micro service is built using the Spring Boot framework and programmed using the Java language.Kafka message queue is selected to realize data transmission between different services,so as to achieve the purpose of asynchronous message processing between services.MySQL and Redis are selected to realize hot and cold data storage and are deployed in a cluster mode.By dividing the MySQL database into databases and tables,the hot data is stored in Redis,and the Nginx reverse proxy server technology is used to distribute the front-end requests to different servers for processing.In the performance preparation micro service,in order to meet the business characteristics of nearby screening self pick-up points,GeoHash address coding algorithm is adopted to improve the efficiency of obtaining the list of stores;In order to optimize the efficiency of store task allocation,a location pool scheduling strategy combined with the actual production scenario is designed to optimize the sorting process,so as to improve the performance of the system and ensure that the system still performs well in the face of a large number of users purchasing fresh goods.The fresh commodity performance system based on micro services has been running smoothly since the cluster was deployed online.The average response time of each module of the test system is about 700 milliseconds,and the average CPU occupancy rate is about 62%.It has realized the full link informatization from consumer orders to final commodity delivery,providing a strong support for the business development of the fresh e-commerce platform. |
