Research And Implementation Of Key Technologies For Live System Cloud Deployment And Personalized Data Processing

In recent years,with the upgrading of Internet technology,the popularization of 4G technology,the 5G era is coming,the audio and video technology is gradually maturing,and Internet video broadcast is increasingly used in various fields of social life,for individual users.Live broadcasts,such as: DOUYU,HUYA live broadcast has been on fire for a while,the live broadcasts application for the enterprise has slowly become popular.Many enterprises and college users need to use live broadcasts for marketing and live academic lectures.The project team developed a live broadcast for the enterprise.Users can start live broadcast on the computer and mobile phone,and can watch the live broadcast through the mobile APP,We Chat applet,computer version website,mobile website,and the live broadcast system also includes the enterprise.The user manages the live broadcast portal subsystem,the super-control console subsystem used by the administrator,and the live portal subsystem.Focusing on this set of live broadcast system,this thesis studies the deployment of the system in the cloud era,the architecture of the system,the reliability and the data generated by the live broadcast.The aim is to improve the user experience of the live broadcast system through the research of these key technologies.To help corporate customers get better service.First of all,this thesis focuses on the live broadcast system for enterprises,and carries out research and implementation of key technologies for cloud deployment,including using NGINX and Keepalived to build a high-availability load balancer for live broadcast system,improving the reliability of the live broadcast system;designing and implementing The front-end and back-end architecture of the entire live broadcast system and the access control of the API are controlled.The research realizes the deployment of the relational database My SQL cluster used in the live broadcast system and the cluster deployment of the nonrelational database Redis,which improves the reliability and security of the data;Based on the Jenkins continuous integration platform research,the automated deployment of the live broadcast system is realized,and the overall efficiency of development,testing and deployment is improved.Secondly,around the live broadcast system,this thesis studies and implements a series of key technologies for personalized data processing.It mainly includes: unified log collection platform based on Filebeat,Logstash,Elasticsearch,Kibana;research and implementation of key technologies of We Chat shared link analysis,analysis of live sharing paths,and promotion of live promotion efficiency;research and implementation of collaborative filtering algorithm Bayesian personalized sorting algorithm for live broadcast recommendation key technology to provide users with a personalized viewing experience;in order to help the live broadcast platform tenants to master the live broadcast related data,this thesis statistics the live broadcast indicator data,including the live broadcast basic profile,live broadcast The audience statistics,the live sub-account statistics,and the PDF reports generated by these statistical indicators,convenient for tenants to download and circulate,improving the user experience.Finally,this thesis introduces the test environment of the system,and tests the functional modules of cloud deployment and personalized data processing to verify the availability of the system.Then test the live broadcast system after cloud deployment,the response time and throughput of the system were tested,and the efficiency and reliability of the system were verified.