Posicionamento de servidores com minimização de latência em redes de operadoras de telecomunicações

Detalhes bibliográficos
Ano de defesa: 2017
Autor(a) principal: Portela, Fábio Affonso
Orientador(a): Não Informado pela instituição
Banca de defesa: Não Informado pela instituição
Tipo de documento: Dissertação
Tipo de acesso: Acesso aberto
Idioma: por
Instituição de defesa: Universidade Federal do Rio de Janeiro
Brasil
Instituto Alberto Luiz Coimbra de Pós-Graduação e Pesquisa de Engenharia
Programa de Pós-Graduação em Engenharia Elétrica
UFRJ
Programa de Pós-Graduação: Não Informado pela instituição
Departamento: Não Informado pela instituição
País: Não Informado pela instituição
Palavras-chave em Português:
Link de acesso: http://hdl.handle.net/11422/6339
Resumo: Telecom service providers are going through a transformation moment, leaving the status of connectivity providers to providers of integrated solutions of telecommunications and information technology (IT). This paradigm shift demands aggregation of process, storage and data management to their network by activating servers in datacenters located at the network. This dissertation studies the deployment of a datacenter in points of presence through the network in order to allow new services. Due to high availability requirements of this kind of service, it is proposed to use geodistribution of servers to increase service survivability. This approach causes increase of latency between servers, which is a factor that could have a negative effect on quality experience of and users. In order to explore the behavior of the commitment between survivability and latency in a geodistributed environment, we formulated an optimization problem with the objective to minimize the average latency between sites. The input parameters are the topology and inter-site latency measurements of an academic research network which provides broadband connectivity to academic facilities. The goal of this dissertation is to analyze the behavior of latency and survivability subject to different topologies using daily and monthly latency values. The results show that there is a band where it is possible to have big earnings in survivability with a little loss of performance in average and maximum latency inter-sites.