Fair: uma ferramenta genérica para migração de dados entre SGBDs

Abstract

The increasing adoption of different database models has made data migration a fundamental activity to ensure interoperability, scalability, and suitability to the specific needs of modern applications. However, the structural differences between relational and non-relational databases make this process complex, requiring mechanisms capable of preserving data integrity and adapting structures between the involved systems. This work presents Fair, a data migration tool designed to operate in a generic and technology-independent manner. The proposed architecture employs an adaptation layer based on the Adapter design pattern and an intermediate model responsible for decoupling source and target systems, allowing greater extensibility and reusability of the migration process. Experiments conducted using the benchmark datasets DVDRental and Chinook 1.4 demonstrated the tool’s ability to perform migrations from PostgreSQL to MongoDB and Neo4j while preserving the structures and relationships of the original data. The results indicated temporal stability in the MongoDB-oriented approach, with migration times remaining close to 5 to 6 seconds even as data volume increased, whereas the Neo4j approach exhibited a progressive growth in processing time.