Distributed Memory Database System Design And Application

With the development of Internet and information technology, especially the global popularization of podcast and increasing growth of web traffic, there are almost a billion daily visitors on the website. So the performance requirement of hardware is becoming increasingly demanding. The bottleneck of database with disk storage is gradually revealed on the performance. The main reason why traditional database system is not suitable for large-scale data process and real-time application is due to its own system architecture. In order to address how to effectively organize large-scale memory data problem, the concept of distributed memory database is introduced. Compared to disk storage, the distributed memory database has much faster data read/write speed by several orders of magnitude. In the meantime, it can implement storage management with dynamic sharing data. So the distributed memory database can improve performance significantly compared to the traditional database which accesses data from disk. Therefore it becomes one of the most important direction and hotspot in the field of modern database research in the recent years.This paper first discusses the related theory of distributed memory database and key technologies, focusing on T-tree index technology which is commonly used in the distributed memory database. In the next place, it introduces some changes to the T-tree index which can improve performance to a certain extent. Once again, it's discussed in details and a simple distributed memory database system is implemented based on the key technologies. This paper has detailed analysis and design for the functionality of each component. Lastly, the validity and high efficient data processing capability of the distributed memory database is verified through system testing.