Research On Distributed Server Architecture Applied To Virtual Reality

With the rapid development of cloud computing and big data,the server cluster often needs to face massive blob(binary large file)data and large concurrency.Therefore,we should solve the amount of business requests and network IO overhead.A large number of virtual reality applications meet the mass use of distributed service cluster and cloud computing resources.In order to meet the requirements of real-time data interaction,high-speed computing,rendering and storage,and so on,and make ensure the security of data,how to design a distributed server architecture to meet the requirements of the virtual reality application for IO intensive services has been a very important lesson.In this paper,the big data distributed computing framework is unable to meet the requirements of virtual reality applications for high real-time services and data intensive services.The research and design of distributed server framework is proposed.This paper analyzes the current research status of distributed systems at home and abroad,and deeply studies and designs distributed server architecture.There are four main parts: the first,optimization of cooperative work and data consistency among cluster nodes;the second is the implementation of distributed database cluster and distributed cache;the third is the design of load balancing algorithm and routing strategy;the last one is the design of the server framework.The key is the implementation strategies and algorithms include network library,collaborative strategy,consistency algorithm,load balancing algorithm,distributed database and algorithm,and node topology.Some distributed servers are introduced in detail.Based on the basic structure and algorithm,some components are adjusted and optimized for the characteristics of virtual reality application and distributed system.In order to verify the architecture proposed in this paper,this paper designs a distributed server cluster system,which is deployed in the Alibaba cloud to provide users with high real-time data services under high concurrency,to meet the requirements of real-time,high availability and high scalability,and to solve the highconcurrency of virtual reality system for high concurrency.The interaction and mass data storage and sending or receiving and provide multiple copies to ensure the high availability and consistency of the system,the server cluster can customize the different strategies according to the specific business.The main development and test platform is the Linux environment.In order to match the cloud computing resources,the running environment is unified.Centos7.2,a total of 10 nodes to carry out the experiment,and to provide users with massive data,data transmission and small and medium computing of 3 types of services,to ensure data security and transmission efficiency.The system ben shows that each module and algorithm of the system designed and implemented can run correctly,and can be deployed in the cloud node to run steadily and provide service in the high and the end to meet the design goal.