Desenvolvimento de um serviço para observabilidade de infraestruturas baseadas em nuvem

Abstract

The expanding reach of technology poses new challenges to developers and system administrators. Problem solving through technological tools, from the most complex to the most trivial, has become quite common. In addition to the value proposition, the infrastructure behind the software is critical, as it enables its full operation so that it can deliver value. Therefore, observability and monitoring are key concepts for successful software. Understanding an application's behaviors and acting proactively before a potential problem occurs are essential to staying competitive in the market. In complex and distributed environments, many variables must be considered when determining the root cause of a problem. Currently, to use the observability tools available on the market, it is necessary to dedicate time and effort to understand the libraries and implement the source code. It is also essential to know performance engineering in order to interpret the metrics and obtain useful information. Given these statements, this work aims to provide a tool to help identify possible bottlenecks and inefficiencies in the adjacent infrastructure of a software, as well as suggest improvements at the machine level based on what was observed. The implemented solution proposes to collect pre-established metrics, analyze the information and bring results that help the system administrator to make decisions, in order to improve its infrastructure.