Arquitetura escalável de alto desempenho para atualização, acesso e recuperação de informações em bancos de dados de aplicações embarcadas

• Ano de defesa: 2012
• Autor(a) principal: Mezzalira, Daniel
• Orientador(a): Trevelin, Luis Carlos
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de São Carlos
• Programa de Pós-Graduação: Programa de Pós-Graduação em Ciência da Computação - PPGCC
• País: BR
• Palavras-chave em Português: Sistemas distribuídos; Arquiteturas distribuídas; Teoria das filas; Escalabilidade; Alto desempenho; Sistema de filas; Simulação; Sistemas embarcados; Telemetria; Redes DTN; Tolerância a falhas
• Palavras-chave em Inglês: Software Distributed Architecture; High performance; Scalability; Queuing systems; Simulation; Embedded systems; Telemetry; DTN; Fault tolerance
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: https://repositorio.ufscar.br/handle/20.500.14289/506

Managing multiple systems such as machine tools, vehicles, aircraft, among others, demand a very intense flow of data between them and the system manager. Researches have been developed in the design and implementation of scalable architectures that meet these demands leading to interesting questions of performance. The objective of this work is to propose a low cost scalable architecture for embedded applications, using pools of personal computers for high performance storage, retrieval and processing of information. It is driven by strong demand for tracking and monitoring of machines and vehicles, covering concepts of mobile networks with satellites and GPRS technology together with the requirement of reliability and performance in the sending of information. Proposes the definition of a server structure, whose distribution is transparent to the application, which is responsible for the receipt of messages from embedded devices via radio frequency technology, decoding and integration of information in the database and subsequent recovery of these information s. Through simulation of different modeled strategies using queuing theory to determine the architecture and the use of predictive mathematical methods for estimating the future burden for the server application, it was possible to obtain a solution that satisfactorily met the assumptions of the research. Thus, it is concluded that it is possible to estimate trends peaks processing information for telemetry applications fleet.