Research And Implementation Of Redis Cloud Database

With the development of network,the business system of all walks of life is becoming larger and larger,and the business content is becoming more and more complex.The general database can no longer meet the requirement of response delay,so Memcached and other caching systems are widely used.The existing cloud platform of the laboratory also uses Memcached as the cache database to improve the opening speed and response speed of cloud applications.However,many problems in high availability,efficiency and persistence are found in the daily development and maintenance.Therefore,based on Redis,this thesis studies and implements a highly available cloud database system called CRDB(CAC Redis Database)as the cache database of the cloud platform.The main work is as follows:Firstly,this thesis analyses the existing problems of the Memcached cached database in the lab and puts forward related solutions.The Memcached used in the current lab is single-point,does not support high availability,and can only process key-value pairs of data.For other complex data types,the client needs to do a lot of data conversion work,which affects the development efficiency.And Memcached does not support persistence,which can be easily lost for some critical data.To solve the above problems,this thesis is based on Redis database engine and proxy-based architecture,proposes CRDB cloud database system,which is mainly composed of access module,proxy module,data pro-cessing module and management module.Then,starting from the four modules of the system,this thesis makes a detailed anal-ysis and design and gives a specific implementation.(1)Access module,as the entrance to the system,mainly handles a large number of requests from the client,uses a combi-nation of LVS and Keepalived architecture for load balancing,which enables the system to handle a large amount of concurrent data,and adds user authentication scenarios for multi tenant access.(2)The proxy module adds multi-process capabilities to meet the high performance requirements of the cloud platform through secondary development of Twemproxy.Using the proxy layer to partition the bottom nodes ensures that the bottom layer can transparently expand the capacity of the upper layer while evenly distributing and forwarding a large number of client requests.(3)Data processing module implements data persistence and uses LRU-K algorithm for Redis's memory replacement strategy to improve the system's processing performance for hot spot data on cloud platforms and improve the application response speed.(4)The management module uses Prometheus to monitor the system and develops various Exporter business plugins to help managers master the system more comprehensively.In addition,the Sentinel architecture is used to complete the function of fault detection and fault handling,which solves the single point problem.And a set of log system is developed based on log4cxx,which provides a good reference for maintenance personnel and developers to solve problems.Finally,this thesis deployed a cloud database on the cloud platform called CRDB,and conducted a comprehensive function and performance test on the CRDB system.Com-paring the result with the original Memcached,the system basically can meet the original design requirements.It also meets the needs of cloud platforms and has good high avail-ability and high concurrency.