Uma abordagem de verificação e validação para sistemas de middleware específicos de domínio dirigidos a modelo

• Autor(a) principal: Fortes, Marcelo Rodrigues
• Data de Publicação: 2018
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFG
• Texto Completo: http://repositorio.bc.ufg.br/tede/handle/tede/9062
• topic: Middleware; Verificação e validação; Engenharia dirigida a modelos; Consistência; Garantia de qualidade; Middleware; Verification and validation; Model-driven engineering; Consistency; Quality assurance; CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO

Middleware platforms aim to facilitate the construction of distributed applications, hiding the complexities and specificities inherent in the underlying platform. However, while middleware facilitates the construction of applications, its own construction is quite complex, requiring good knowledge in software design and implementation. In this sense, some approaches have been proposed, with the aim of presenting more flexible and configurable ways to build middleware, for example, reflective middleware, model-driven middleware and component-based middleware. Another example is the Model-Driven Domain-Specific Middleware (MD-DSM) approach, which employs Model-Driven Engineering concepts for building middleware platforms that support the execution of model-based applications. MD-DSM solves several problems related to building middleware for different domains. However, the entire process of verifying and validating the final MDDSM product is performed in an ad hoc way, without a methodology that guides the middleware engineer during these activities, possibly compromising the quality of the final product. In this work, we present a verification and validation (V&V) methodology that systematizes the entire quality assurance activities of this category of middleware. In addition, we present a tool that automates much of the V&V activities that need to be performed to ensure the quality of a middleware system built using MD-DSM. We also present a quantitative evaluation of the V&V tool.