This repository can be used to create a monitoring stack based on the following components:
- Prometheus
- Alertmanager
- Pushgateway
- Grafana
- Alerta
- Karma
There's a docker-compose that can be used to deploy every component using docker.
All the services on the docker-compose have health-checks.
Prometheus configuration collects metrics from Alertmanager, Karma, Pushgateway and Grafana using static config.
If you want to add targets from node_exporter or telegraf you can create the following files under ./prometheus:
- ./prometheus/node_exporter-targets.json
- ./prometheus/telegraf-targets.json
Rules are under ./prometheus/rules.
There's a file prometheus-consul.yml that uses Consul has a Service Discovery and also creates job name.
For this to be done automatically you just need to register the service on Consul and add the prometheus tag.
Prometheus is listening on port 9090.
The default configuration sends alerts using webhooks to Alerta and Karma gets also alerts from Alertmanager automatically.
For Slack integration change the alertmanager/alertmanager.yml file.
Alertmanager is listening on port 9093.
The Pushgateway has the default configuration.
Pushgateway is listening on port 9091.
Grafana is provisioned with datasource for Prometheus and some dashboards.
Grafana is listening on port 3000.
Default configuration with MongoDB has database.
Alerta is listening on port 8088.
Default configuration gets alerts from Alertmanager.
Karma is listening on port 8080.
If you want to have LTS install M3DB, uncomment prometheus.yml configuration and change the M3DB server.
- Alerta health-check
- MongoDB health-check
- Alerta prometheus metrics