Converting Relational Database to Document-Oriented NoSQL Cloud Database

Abstract

In the Internet of Things (IoT) area, the increase of data leads to the "Big data'' problem. The traditional relational database (RDB) is not being able to deal with processing big data. In contrast, the Not only SQL (NoSQL) database was created to deal with big data problems. Therefore, most organizations need to convert their data stored in RDB systems into NoSQL using flexible models and processes. Most research into the conversion of RDBs into newer database systems has concentrated on schemata transformation, RDB data publishing, whereas other little work focuses on data migration. This paper proposes an automatic approach to convert a database implemented in MySQL database management system into an equivalent database for the target MongoDB system, which is a leading NoSQL database system. This method can handle a large amount of existing data in RDBs without any loss or changes in data semantics and data instances. The solution takes an existing RDB as input, extracts its schema, and then analyzes the schema in an array to be converted with data instances according to the NoSQL target database structure. A system of the proposed approach has been designed. An experimental study was performed to evaluate the proposed approach. The experimental results show that the target database generated by the prototype and the target database generated by other methods were comparable and equivalent.