Liveness probe
This page explains how to configure the OpenCTI liveness probe, a lightweight HTTP endpoint that reports whether the platform process is running.
What is the liveness probe?
The liveness probe is a dedicated HTTP server that starts immediately when the OpenCTI process launches, before any platform initialization (database migrations, dependency checks, cache warm-up, etc.). It responds with HTTP 200 OK to indicate that the process is alive.
This is distinct from the existing /health readiness endpoint, which only becomes available after the full platform has started and checks all backend dependencies (ElasticSearch, Redis, RabbitMQ, S3).
Why use it?
Container orchestrators like Kubernetes need to distinguish between two states:
- Liveness — Is the process running? If not, restart the container.
- Readiness — Is the application ready to serve traffic? If not, stop sending requests.
| Probe | Endpoint | Port | Available | Checks |
|---|---|---|---|---|
| Liveness | {base_path}/health/liveness |
app:liveness_port (default disabled) |
Immediately on process start | Process is running |
| Readiness | {base_path}/health |
app:port (default 4000) |
After full platform initialization | ElasticSearch, Redis, RabbitMQ, S3 |
What's next?
- Configuration — Full list of platform configuration parameters.
- Clustering — Deploy multiple OpenCTI instances with health monitoring.
- Troubleshooting — Diagnose startup and connectivity issues.