MigDB - relational to NoSQL mapper

摘要

Relational databases have been the foremost means of database management over the years. Relational databases, which provide vast benefits like the operational ease and sound execution of transactions with its support for ACID properties, are ideal when it comes to structured data. Nonetheless, with regards to scalability and the emergence of big data which encompasses escalating amount of structured and unstructured data, relational databases fall short in becoming a proper fit. Addressing these concerns, NoSQL comes in as a much more flexible alternative to relational databases. With the popularity of NoSQL which is capable of handling vast amounts of complex data varying in type and structure, the focus moves on the migration of existing relational databases into NoSQL databases, especially databases like MongoDB. In consideration of the migration of relational database tables, data along with queries into MongoDB, there are certain issues like current migration systems being incapable of properly mapping table relationships. It could result in the necessity of multiple query statements, thereby decreasing query performance and data redundancy thus wasting space needlessly. Adding up to those complications are the existing query converters only being capable of converting select queries, the inability of query converters to support table joins and most of the tools requiring MongoDB command or JSON knowledge. Hence, the focus of the intended system `MigDB' is to facilitate users to easily transfer from MySQL tables along with proper relationship mapping opted from embedding and referencing, data as well as queries into MongoDB without requiring any prior knowledge of MongoDB commands or JSON.