Setting up a monitoring system for the analysis and monitoring of docker containers. In order to be able to clearly display the resource consumption and uptime of docker containers, collected metrics of both the container and the docker host are displayed in a dashboard.
Supplement
All components used for monitoring run in docker containers. Monitoring consists of two docker stacks: An exporter stack consisting of cAdvisor and Node Exporter that reads container (cAdvisor) and host (Node Exporter) metrics such as memory and CPU utilization. The second stack consists of Grafana for visualizing the data and two time series databases, Prometheus and Influxdb. Prometheus for storing (short-lived) monitoring data and Influxdb for persisting long-term analysis data.
Subject description
Metric exporters (cAdvisor, node exporters) query this data and make it available via interfaces. The Prometheus database reads the provided data and saves it. For a long-term analysis of the data, an Influxdb database is available, which downsamples the data and is filled by Prometheus. Grafana serves as a web-based visualization tool for the incoming metrics