SILVA, M. M.; http://lattes.cnpq.br/7478582255896370; SILVA, Mariana Mendes e.
Resumen:
The IaaS model brings many challenges, one of which is the dynamic resource provisioning, which is made by an auto-scaling mechanism. Configuring and evaluating this mechanism is a complex task. There are many works about it in the literature, but most are about new algorithms. Only a few consider configuration and evaluation and don’t emphasize the practical aspect, considering the difficulties that probably exist in the industry context. To better understand, we built a case study in a company client from a cloud provider. We wrote a questionnaire e analyzed the results. We created a tool to collect data to evaluate infrastructure utilization. We understood the most significant difficulties in configuration and evaluation with these two tools - questionnaire and collector - and then correlated them with infrastructure data analysis. In the end, we also made a mapping with improvement points. Therefore, this research takes the first steps towards systematizing the evaluation and practical configuration of an auto-scaling service.