upSince, and uptime — making it ideal for monitoring, load balancer probes, and orchestration systems.

Enable the health_check extension
Add thehealth_check extension to your collector configuration and bind it to the default address and port (0.0.0.0:13133):
Verify the endpoint with curl
A GET request to/ returns HTTP 200 and a JSON body describing server availability:
The extension listens by default on port 13133. You can use this endpoint in health checks, load balancer probes, or external orchestration tools to decide rollout and restart behavior.
Kubernetes: map to liveness and readiness probes
Use the health endpoint for Kubernetes liveness and readiness checks so the platform can detect and recover from unhealthy collector instances:| Probe type | Purpose | Example |
|---|---|---|
| Liveness | Restart container when collector is unresponsive | livenessProbe YAML above |
| Readiness | Exclude pod from service load balancer while starting/rehydrating | readinessProbe YAML above |
Docker / Docker Compose
For Docker or Docker Compose deployments, expose the collector port so external systems (host, orchestrator, CI) can poll the endpoint:http://<host>:13133/ for a quick health signal and use the 200 OK response for rollout, restart, or monitoring decisions.
When to use the healthcheck extension
- Integrate with Kubernetes probes (liveness & readiness).
- Configure load balancer health checks to avoid routing traffic to an unavailable collector.
- Use in CI/CD pipelines or deployment scripts to gate rollouts on collector availability.
- Provide a simple browser-accessible endpoint for quick diagnostic checks.
Links and references
- OpenTelemetry Collector — Extensions
- Kubernetes Probes — Liveness and Readiness
- Docker Compose — Expose ports