CloudMetric: um arcabouço para criação e monitoramento de métricas personalizadas em nuvens computacionais

• Ano de defesa: 2019
• Autor(a) principal: Santos, Pablo Brunetti dos
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal do Espírito Santo; BR; Mestrado em Informática; Centro Tecnológico; UFES; Programa de Pós-Graduação em Informática
• Palavras-chave em Português: Monitoramento; OpenStack; Nuvens computacionais; Monitoring; Computational clouds; subject.br-rjbn; Ciência da Computação
• Link de acesso: http://repositorio.ufes.br/handle/10/13745

The computational clouds need constant measurement and monitoring to ensure the proper functioning of the different applications hosted there. Questions such as “ How to monitor? ”, “ What to monitor? ” And “ How to integrate application performance metrics with the cloud? ” Are relevant aspects that need to be investigated. However some cloud platforms do not offer a customizable resource monitoring service and means of creating specialized metrics for different resources and applications. All metrics creation and monitoring is performed in a script and decentralized manner, ie there is no possibility of integration between multiple OpenStack clouds, for example. Therefore, this paper proposes a framework for the creation and monitoring of custom metrics in computational clouds. To validate CloudMetric, a prototype based on the OpenStack cloud computing platform was built, which allows monitoring of metrics created and managed by the Ceilometer module and the creation and monitoring of custom metrics by CloudMetric itself. Monitoring is performed by querying metrics stored in the Gnocchi temporal database, viewing the metrics in the Grafana dashboard viewer builder and exporting to a file in json format. Through this framework it was demonstrated that CloudMetric was able to perform metric creation in a standardized manner without using the OpenStack native programming APIs with fast, simple and robust monitoring. A demonstration of CloudMetric’s functionality is presented in 2 test cases: a robot’s path control in an intelligent space and web server metric monitoring. In addition, tutorial documentation allows the use of this framework in future computational cloud-based research projects