Research And Implementation Of Cluster Storage System Based On Linux Environment

In the 21 st century,people entered the era of digital information,and various electronic products came into being.The massive data storage problems of schools,enterprises,and government agencies have become imminent.According to a reliable survey,the total amount of global data in 2012 reached 2.8ZB,and it is expected to reach 40 ZB by 2020.The storage and backup of big data has become the direction that every company strives to pursue.The necessity of cluster storage is already very obvious.It can not only provide enterprises with the convenience of data storage,but also provide guarantee for the country to develop higher-end technology products.The national science and technology level reflects the country's overall strength,and it is of great engineering significance to study and improve the cluster storage platform to solve the problem of excessive data in the process of research and development of new technology products.The big data cluster storage system studied in this paper is a distributed data processing system.A clustered storage management system based on Linux operating system is designed and applied in practical projects.The main research contents of the thesis are divided into the following four parts:(1)According to the research status of cluster systems at home and abroad,the complete architecture of the whole cluster system is carried out,the engineering application scenarios are roughly introduced,and various open source components and interfaces are introduced in detail.(2)Node monitoring function design: In order to monitor the working status of the entire cluster system node in real time,set the management server to check the cluster host condition at regular intervals,including network read and write rate distribution,disk occupancy rate distribution,memory occupancy rate distribution,and CPU Occupancy distribution;(3)Focusing on the most critical load balancing technology in the entire cluster system,focusing on the phenomenon of overloaded nodes in large-scale distributed clusters at a certain time,the existing load balancing algorithm is optimized to makeThe cluster server can distribute the existing data evenly to improve the service life of the server.(4)Analyze the currently available high-availability solutions in the cluster system,and design high-availability cluster solutions that meet the project requirements in practical engineering applications,and test and adjust them.The research and test show that the design and research results of the cluster system discussed in this paper can be well applied to the actual project,and in practical applications,the cluster system can be upgraded in real time according to the needs of the enterprise.