SCPNET - Uma arquitetura de monitoramento de rede baseada em políticas

• Ano de defesa: 2018
• Autor(a) principal: Silva, Nilton Camargo Batista da
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; Brasil; Ciência da Computação; UFSM; Programa de Pós-Graduação em Ciência da Computação; Centro de Tecnologia
• Palavras-chave em Português: Monitoramento de redes; Gerenciamento baseado em política; PBNM; Big data; Correlação de dados; Network monitoring; Policy-based management; Data correlation; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: http://repositorio.ufsm.br/handle/1/14060

Computer networks are constantly growing, which causes a huge increase in the number of data and difficult their management. In this context, there is a recent interest in applying Big Data concepts, tools and technologies in network management. Along with the large volume of data, one of the limitations of network management concerns the heterogeneity of technologies used in modern computer networks. Therefore, for the administrator to achieve a uniform management of networks, it is necessary that the management system used would be able to interpret objectives in an abstract way. PBNM is a technique that uses the concept of policies so that the network administrator can define what are accomplishments to be achieved, using a more abstract language. Nevertheless, to date, the use of such techniques in the context of network monitoring remains unknown. Thus, the main objectives of this dissertation are to propose solutions which would allow the network administrator to express abstractly the devices and services that one wants to be monitored in an environment capable of handling large numbers of data. In order to achieve these objectives, an architecture is proposed in order to allow the network administrator to express, in an abstract way, the devices and services that are wanted to be monitored in an environment capable of handling large numbers of data. Based on this architecture, a prototype was developed. Finally, the prototype developed, along with the policy model for network monitoring, was evaluated in a test environment, where it was possible to perceive some advantages when using policies while monitoring networks, as there was there was no overload in the request-response messaging network. The results indicate that the use of policies together with data correlation in a Big Data environment can be used as a mechanism for monitoring networks.