EC
Liveness probe failed
KubernetesERRORNotableContainer ErrorHIGH confidence

Container killed by failing liveness probe

Production Risk

Repeated container kills cause service interruptions and increased RESTARTS count.

What this means

When the liveness probe fails consecutively beyond failureThreshold, Kubernetes kills and restarts the container. The liveness probe is intended to detect a deadlocked or unrecoverable application state. However, misconfigured liveness probes (too sensitive thresholds, wrong endpoint) can cause unnecessary container kills and service disruption.

Why it happens
  1. 1Application entered a deadlocked state and cannot respond to the liveness endpoint
  2. 2Liveness probe endpoint path or port is misconfigured
  3. 3timeoutSeconds or failureThreshold set too low causing probe to fail during normal load spikes
  4. 4Downstream dependency failure causes the liveness endpoint to return errors
How to reproduce

Container restarts repeatedly; events show liveness probe failures.

trigger — this will error
trigger — this will error
kubectl describe pod mypod | grep -A 5 "Events:"
# Warning  Unhealthy  kubelet  Liveness probe failed: HTTP probe failed with
# statuscode: 500

kubectl get events --field-selector involvedObject.name=mypod,reason=Unhealthy

expected output

Warning  Unhealthy  ...  Liveness probe failed: HTTP probe failed with statuscode: 500

Fix 1

Test liveness endpoint from inside the container

WHEN Probe may be hitting the wrong path or port

Test liveness endpoint from inside the container
kubectl exec mypod -- wget -qO- http://localhost:8080/healthz
kubectl describe pod mypod | grep -A 10 "Liveness:"

Why this works

Verifies the probe endpoint is accessible and returning a healthy response.

Fix 2

Tune probe thresholds to avoid false positives

WHEN Probe fails during legitimate load spikes

Tune probe thresholds to avoid false positives
livenessProbe:
  httpGet:
    path: /healthz
    port: 8080
  initialDelaySeconds: 30
  periodSeconds: 15
  timeoutSeconds: 5
  failureThreshold: 5

Why this works

Higher failureThreshold and timeoutSeconds tolerate transient slowness without killing the container.

What not to do

✕

Sources
Official documentation ↗

Kubernetes Documentation

Content generated with AI assistance and reviewed for accuracy. Found an error? hello@errcodes.dev

← All Kubernetes errors