Uma arquitetura autonômica para a alocação de recursos através de migração de serviços em ambientes Fog Computing

• Ano de defesa: 2018
• Autor(a) principal: Silva, Danilo Souza
• Orientador(a): Ribeiro, Admilson de Ribamar Lima
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Programa de Pós-Graduação: Pós-Graduação em Ciência da Computação
• Palavras-chave em Português: Computação autonômica; Sistemas distribuídos; Provisionamento de recursos; Internet das Coisas
• Palavras-chave em Inglês: Fog computing; Autonomic computing; Distributed systems; Resource provisioning; Internet of Things (IoT)
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: http://ri.ufs.br/jspui/handle/riufs/11229

In recent years, the number of smart devices (eg smartphones, sensors, autonomous vehicles) has grown significantly. In this scenario, the computational demand needed to meet the demand for latency-sensitive applications in domains such as IoT, Industria 4.0 and smart cities is also growing and the traditional cloud computing model is no longer able to meet all of its needs alone of application. As an alternative to this limitation, a new computing paradigm called Fog Computing was introduced. This paradigm defines the architecture that extends the computational capacity and storage of the cloud to the edge of the network. However, one of the main problems is how to efficiently determine where services will be allocated to meet certain QoS requirements for provisioning services through IoT applications. The present study aims to present an optimization strategy for the resource allocation problem using service migration through an autonomic architecture model based on the MAPE-K control loop. Based on the presented model, the strategy was implemented with container virtualization technology and evaluated in a large scale virtual environment for IoT, called VIoLET. The results show that it is possible to optimize a fog computing environment using the service migration between the nodes according to established objectives and autonomously. The work contributes with a bibliographical review of the state of the art on resource management, the implementation of a monitoring and orchestration environment for VIoLET, as well as contributing to the development and evaluation of the optimization strategy as well as the analysis of resource utilization of the proposed solution. Finally, we conclude the paper by presenting a list of promising research directions outlined for future work. We expect the work to serve as a basis for research that seeks to develop optimization techniques for resource utilization in Fog Computing environments.