Análise comparativa de desempenho entre procedimentos armazenados em SQL e processamento de dados em lote em softwares corporativos

Abstract

The growth of the availability of information is accompanied by an increasing diversity of methods for data processing. Traditionally some of these methods are well-established; however, some may offer better performance compared to others in specific scenarios. A comparative analysis between methods can be useful to support choices in the context of corporate software programming, which by its nature makes constant use of data processing. This work compares data processing through the batch processing method versus processing using stored procedures in SQL. Using tools such as PostgreSQL as the Database Management System (DBMS), an ERP simulator in Python, native performance analysis tools, and other tools for data generation, this study conducts a comparative performance analysis between these data processing methods in the context of corporate systems. The study concluded that although one processing method shows better performance in terms of execution time and memory usage, conceptual factors inherent to each approach, such as ease of maintenance, scalability, security, system integration capability, among other factors, are fundamental for comparison.