Design of Large Data Storage Structure of Database Based on Data Mining

Abstract

In view of the development of information construction but the performance level of business database is relatively backward, this paper puts forward a solution to solve the problem of large amount of data and low efficiency of database in the process of system database design. Based on data mining matching algorithm, the model classifies data by analyzing data flow, data access characteristics, data volume, data growth and data life cycle, so as to improve the performance of database. Based on MySQL sharding technology, an efficient and scalable storage architecture is designed. This paper proposes a three-layer splitting model, which is divided into database, horizontal table and partition, and uses various solutions, such as vertical and horizontal segmentation, to hash data into different databases or relational tables. At the same time, the distributed data access layer is designed for ORM framework, which effectively relieves the pressure on the database and improves the efficiency of data access.