FLX_APP_PROMETHEUS_003 Alert
This alert is triggered when one or more exporters are down in a node. Exporters are metric collecting agent which runs on all flex nodes.
When the exporter is down, metrics are not gathered and will not be exported to metric servers like Prometheus. This will result in no metric data of the affected node( no graph in grafana). Metric usage is vital as it shows resource usage statistics for each node and is also used to monitor resource usage for each node and raise alerts accordingly. The server is unavailable to the monitoring server without the metrics exporters running.
Login to the nodes where the exporter down error is reported.
$ ssh SERVER
Note the health state of the affected exporter.
$ docker ps | grep <exporter name>
You need to restart the service if the health status is 'unhealthy' or 'down'
$ docker compose stop <exporter name>
$ docker compose start <exporter name>
Check the health status after restarting.
$ docker ps | grep <exporter name>
Check the logs of the exporter services
$ docker logs <exporter name>