Detalhes bibliográficos
| Ano de defesa: |
2019 |
| Autor(a) principal: |
YADAV, Rajeev Ranjan
 |
| Orientador(a): |
SOUSA, Erica Teixeira Gomes de |
| Banca de defesa: |
LINS, Fernando Antonio Aires,
TAVARES, Eduardo Antonio Guimarães |
| Tipo de documento: |
Dissertação
|
| Tipo de acesso: |
Acesso aberto |
| Idioma: |
por |
| Instituição de defesa: |
Universidade Federal Rural de Pernambuco
|
| Programa de Pós-Graduação: |
Programa de Pós-Graduação em Informática Aplicada
|
| Departamento: |
Departamento de Estatística e Informática
|
| País: |
Brasil
|
| Palavras-chave em Português: |
|
| Área do conhecimento CNPq: |
|
| Link de acesso: |
http://www.tede2.ufrpe.br:8080/tede2/handle/tede2/8556
|
Resumo: |
Cloud computing is emerging as the main mechanism for processing large amounts of data. In this context, private clouds provide efficient infrastructures that support the analysis of data sets generated by different sources, such as social networks, health data, and climatology data. Understanding how Big Data parsing behaves in private cloud environments is an important approach to identifying critical performance and cost factors in these environments. Performance and cost evaluation provides support to manage these environments by considering performance metrics such as processor and memory utilization of virtual machines, and cost metrics such as infrastructure cost, power consumption cost and software cost of these environments. This paper presents a strategy based on a methodology and models to evaluate Big Data transactions supported by a pool of resources provided by the infrastructure of the private cloud. A methodology is proposed to evaluate the performance and cost of Big Data environments in private clouds. This methodology presents activities such as understanding and configuring the Big Data environment in the private cloud, the design of experiments, performance, and energy consumption measurement, performance modeling and cost modeling. A performance model is based on stochastic Petri nets is proposed to estimate resources utilization of virtual machines and cost models consider the cost of deploying a private cloud, costs associated with the energy consumption of the data set analysis, and costs related to the acquisition of related software. The case study illustrates the applicability of the methodology, performance model, and cost models in a real private cloud and provides important information about these topics, such as identifying factors that most impact processor utilization and virtual machine memory, and the energy consumption in these environments. The case study considered the analysis of a data set composed by opinions of the Twitter social network users regarding the 2018 Brazilian’s presidential election. |
