A carregar...
Publicação
Usage Statistics
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 626 KB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
By looking at logs, metrics and traces, it is possible to infer what is going on inside a
system, in order to detect problems or inefficiencies. Quidgest is a global technological
company and since its establishment in 1988, it has pioneered the use of AI applied to
modelling and automatic generation of software. Genio is a tool that allows functional
specialists and analysts to build and support information systems. Quidgest needs to evaluate the performance of the systems generated by Genio. The goal of this work is to
integrate a dashboard with usage and performance statistics into the administration interface of the solutions generated by Genio. To generate mock metrics QuidServer is used, a
windows service developed and used by Quidgest. This service reads and configures long
duration processes and calls external services that run those processes. With this, an event
collection agent was developed in C Sharp. This agent reads events in real time, parses
them and writes them into an InfluxDB bucket. With InfluxDB, it is possible to create continuous queries that run automatically and periodically to downsample the data as needed.
Grafana is then used to create dashboard, which allows for simultaneous visualization of
different data. Having the processing time of received messages as a metric implemented
by the company, a new metric is implemented, the number of invoked scheduling tasks,
as well as the processing time of said tasks. This is done to understand how metrics are
collected from a system: measuring data and aggregating it into metrics that can be sent
through ETW events to be captured by the metric collection agent. Finally, Docker is used
to run InfluxDB in one container and Grafana in another, which allows for the automation
of the installation and configuration of InfluxDB and Grafana.
Descrição
Trabalho de projeto de mestrado, Informática, 2022, Universidade de Lisboa, Faculdade de Ciências
Palavras-chave
TSDB InfluxDB Dashboard Grafana Docker Trabalhos de projeto de mestrado - 2023