Menu
Open source

How to monitor controller

The Grafana Agent Flow component controller exposes Prometheus metrics which you can use to investigate the controller state.

Metrics for the controller are exposed at the /metrics HTTP endpoint of the Grafana Agent Flow HTTP server, which defaults to listening on http://localhost:12345.

The documentation for the grafana-agent run command describes how to modify the address Grafana Agent Flow listens on for HTTP traffic.

The controller exposes the following metrics:

  • agent_component_controller_evaluating (Gauge): Set to 1 whenever the component controller is currently evaluating components. This value may be misrepresented depending on how fast evaluations complete or how often evaluations occur.
  • agent_component_controller_running_components (Gauge): The current number of running components by health. The health is represented in the health_type label.
  • agent_component_evaluation_seconds (Histogram): The time it takes to evaluate components after one of their dependencies is updated.
  • agent_component_dependencies_wait_seconds (Histogram): Time spent by components waiting to be evaluated after one of their dependencies is updated.
  • agent_component_evaluation_queue_size (Gauge): The current number of component evaluations waiting to be performed.