K8s livenessProbe and readinessProbe for worker pod

[L_S]

Is there any recommended way to set up livenessProbe and readinessProbe for worker pods? Worker is not responding to http probes, port 5678 is closed.

[MutedJam]

Hi @L_S, welcome to the community!

You would need to set the QUEUE_HEALTH_CHECK_ACTIVE environment variable to enable the /healthz endpoint.

[pemontto]

See QUEUE_HEALTH_CHECK_ACTIVE env var and use the /healthz endpoint to check

Environment variables | n8n Docs

Worker Health Check - k8s deployment

Hi, I’m currently trying to find a way to provide a health check for the workers, so that unhealthy workers get automatically shutted down by the k8s cluster. Unlike the main node and the webhook nodes, there’s no exposed HTTP endpoint for a worker. I also can’t find any obvious commands that I could execute in the container to check if it is healthy. Any suggestions here? Thanks!

[L_S]

Thank you, @MutedJam and @pemontto. I missed it out in docs. It is working like a charm!

Containers:
  n8n-worker:
    Container ID:  docker://ccb40f7df19fc03631f4ed26bad92563d2a9e49544276eae29f1c8e9c3c81ae6
    Image:         n8nio/n8n
    Image ID:      docker-pullable://n8nio/n8n@sha256:46482af570333c5110225daebfae850f4591764ace4428433975ee6223a36033
    Port:          5678/TCP
    Host Port:     0/TCP
    Command:
      n8n
      worker
    State:          Running
      Started:      Tue, 25 Oct 2022 09:17:06 +0200
    Ready:          True
    Restart Count:  0
    Liveness:       http-get http://:5678/healthz delay=0s timeout=1s period=10s #success=1 #failure=3
    Readiness:      http-get http://:5678/healthz delay=0s timeout=1s period=10s #success=1 #failure=3