Análise comparativa de linguagens e frameworks para implementações de backend

• Autor(a) principal: Pinto, Bruno Augusto Pereira
• Data de Publicação: 2023
• Tipo de documento: Trabalho de conclusão de curso
• Idioma: por
• Título da fonte: Repositório Institucional da Universidade Federal do Ceará (UFC)
• Texto Completo: http://repositorio.ufc.br/handle/riufc/81746
• topic: CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO; Backend; Consumo de CPU; Consumo de memória; Node.js; Javascript; Python; Flask; Backend; CPU consumption; Memory consumption; Node.js; Javascript; Python; Flask

Given the use of computational metrics to price the infrastructure usage in cloud computing, a thorough understanding of how languages and frameworks function proves to be crucial. In doing so, this work aims to comprehend the strengths and weaknesses of the two primary backend development technologies in the market. This study conducts a comparative analysis between Python and Node.js backend implementations, which are the two most widely used programming languages according to the Stack Overflow survey of the current year. The research encompasses various aspects of performance, including CPU and memory consumption, under varying workload scenarios. To achieve this, an identical CRUD implementation was carried out in both languages and subjected to different load volumes. The results reveal interesting nuances in the behavior of the implementations, highlighting relative efficiency in different contexts. Regarding CPU consumption, the Node.js implementation consistently proved to be more efficient in light workload scenarios, indicating superior performance in quick and less intensive operations. On the other hand, the analysis of memory consumption showed that both languages can maintain stable levels of efficiency, adapting remarkably to variable workloads. The Python implementation exhibited a slight reduction in consumption, suggesting effective adaptation to more intense loads, while the Node.js implementation maintained efficient levels, albeit with a more pronounced variation. This work contributes to understanding the performance nuances between Python and Node.js, fostering the development of more effective and efficient solutions in an ever-evolving scenario.