Question

A Java application Deployment shows its pod restarting every couple of minutes. The container starts fine but takes about 45 seconds to fully initialize before it can serve traffic. The pod spec includes a liveness probe: httpGet on /healthz port 8080, initialDelaySeconds: 10, periodSeconds: 10, failureThreshold: 3. Application logs show the process being terminated mid-startup with no application errors. What is the MOST likely cause and the correct fix?

Accepted Answer

The liveness probe begins checking before the app has initialized, so kubelet kills the container; increase initialDelaySeconds (or add a startupProbe) to allow for the ~45s startup time

Answer

The container is being OOMKilled during startup; increase the memory limit in the pod spec

Answer

The readiness probe is failing, which causes kubelet to restart the container; remove the readiness probe

Answer

The image is corrupt and failing to pull; delete the pod so it re-pulls a clean image layer