Skip to main content

Monitoring

The Operator Node can be configured to send Opentelemetry metrics, logs and traces to an OpenTelemetry Collector. The OpenTelemetry Collector can then be configured to send the metrics to a variety of backends such as Prometheus, Grafana, or other monitoring tools.

To configure the Operator Node to send metrics to an OpenTelemetry Collector, you can use the following CLI arguments:

drosera-operator node --otel-export-endpoint <endpoint> --otel-export-metadata <metadata> --otel-resource-attributes <attributes>

Description of the CLI arguments can be found on the configuration section of the Run the Node page.

Logging​

If running the Operator Node as a systemd service, all logs will be persisted on the machine and will consume storage space. To avoid this, you can configure the otel export endpoint to send logs to a remote logging service and send the stdout logs to /dev/null when starting the the Operator Node. This will prevent logs from being stored on the machine.

drosera-operator node --otel-export-endpoint <endpoint> --log-output stdout > /dev/null

Metrics​

The Operator Node collects the following metrics:

System Metrics​

  • drosera_process_cpu_usage: The CPU usage of the Operator Node process
  • drosera_process_disk_space_usage: The disk space usage of the Operator Node process
  • drosera_process_memory_usage: The memory usage by the Operator Node process
  • total_memory: The total memory available on the system

Drosera Metrics​

  • execute_trap_duration: The duration of time it takes to execute a Trap
  • attestation_consensus_duration: The duration of time it takes to reach consensus on an attestation of a Trap result
note

More metrics will be added in the future.