Skip to main content
Use this page after the node starts cleanly. It covers the endpoints and telemetry knobs you can wire into probes and observability.

​
Health

Use /healthz for load balancers, container probes, and uptime checks.

​
Metrics

Enable the Prometheus server with telemetry and use the metrics port for scraping.

​
Logging

Switch to Json when you want structured logs for a collector or log shipper.

​
Tracing

Enable tracing under [log.tracing]. The current node builds its OTLP gRPC exporter with the library’s default collector endpoint; TinyCloud config does not currently expose an endpoint override.
The node exports OTLP spans; your collector can forward them to whatever backend you use when it is reachable at that endpoint.

​
What to avoid

  • Do not expose the metrics port publicly
  • Do not rely on tracing being on by default
  • Do not assume the Prometheus server starts unless telemetry is enabled
The node’s tracing code uses OpenTelemetry OTLP export. The docs here stay at that level instead of naming a specific collector backend.