Migrating from SQL to NOSQL Database: Practices and Analysis

Big data and data analytics require migrating from relational databases (SQL) to NoSQL data structures to represent the data. Such transformation is challenging because of the lack of automatic transformation process and the requirement of guaranteeing both performance and accurate representation. In this paper, we analyze and implement commonly used mapping from SQL to NoSQL structures. We compare between these mappings in terms of retrieval time in an attempt to identify the best mapping. We use MySQL as DBMS for SQL structure and MongoDB for NoSQL structures. Our experiments showed promising results when using a mix of one level embedded document with a reference relationship with another document.