Research And Design Of Storage System Based On Improved Raft

Traditional storage faces challenges: data is growing exponentially,costs are increasing,and it is difficult to manage.With the innovation and development of new technologies,Internet applications are gradually transforming to a distributed system that operates harmoniously among services and provides services uniformly.The advantage of distributed system is that it can make full use of the computing power of multiple machines to build large and complex applications.In the implementation of a distributed system,many problems need to be solved,one of which is how to ensure the reliable availability of data storage in the system.Distributed storage is a technology that uses different storage servers to share the storage burden on different independent devices.The distributed storage system has better accessibility and larger capacity,and is more suitable for large data performance requirements.this thesis analyzes the shortcomings of traditional raft algorithm,such as low log replication efficiency and long election cycle,and proposes an improved raft algorithm based on proxy nodes.this thesis analyzes the problems with the original traditional deployment of memcached.These include downtime,data loss,and load imbalance.Based on the traditional deployment of memcached,this thesis proposes a RDB distributed storage system using the raft consensus algorithm.RDB distributed storage system includes load balancing module,network interface module,system control module,storage control module.The load balancing module adopts keep Alive+HAProxy technology to deploy,and adopts static and static load balancing algorithm.The network interface module uses primary and secondary Reactor threads.The system control module realizes fault migration and consistency.The fault migration mechanism depends on subjective and objective logoff mechanisms.The implementation of consistency mainly relies on raft consistency algorithm.Storage control module implements data storage and logging.This thesis tests the performance of both the traditional memcached and RDB deployments,and the results show that the RDB deployments outperform the memcacheed deployments in both load balancing and hit ratio.