kubernetes networking_error ai_generated true

Readiness probe failed: endpoint not ready, no endpoints available for service

ID: kubernetes/k8s-endpoint-not-ready

Also available as: JSON · Markdown
87%Fix Rate
88%Confidence
60Evidence
2023-01-01First Seen

Version Compatibility

VersionStatusIntroducedDeprecatedNotes
8 active — — —

Root Cause

Service endpoints are not ready because pods haven't passed readiness probes, selectors don't match, or pods are in wrong namespace.

generic

Workarounds

  1. 92% success Verify selector labels match between Service and Pod: kubectl describe service <name> 
    Compare service selector with pod labels: kubectl get pods --show-labels | grep <selector-value>

    Sources: https://kubernetes.io/docs/concepts/services-networking/service/#defining-a-service

  2. 88% success Fix readiness probe configuration so pods can pass the health check 
    kubectl describe pod <name> to check readiness probe failures, then fix the probe path/port/timeout

    Sources: https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/

Dead Ends

Common approaches that don't work:

  1. Remove readiness probes to make endpoints always ready 80% fail

    Sends traffic to unready pods — causes application errors and timeouts

  2. Restart the kube-proxy pods 75% fail

    kube-proxy reads from endpoints API — if endpoints aren't ready, restarting kube-proxy won't help

Error Chain

Leads to:
Preceded by:
Frequently confused with: