How to fix Pod grace period exceeded, force killing pod in Kubernetes

What this error means

Kubernetes sends SIGTERM to pods during shutdown (termination). The grace period (default 30 seconds) is the time allowed for graceful shutdown. If the pod does not terminate within this window, Kubernetes force-kills it (SIGKILL). This error indicates your application is not handling SIGTERM properly or takes longer than the grace period to shut down cleanly.

Typical symptoms

• Pods show "Terminating" status for extended periods
• Error logs: "Pod grace period exceeded, force killing"
• Application logs show no graceful shutdown sequence
• Kubernetes events show force deletion warnings
• Database connections lost abruptly without cleanup
• Deployment rollouts stall during termination
• Job pods terminate ungracefully without saving state

Common causes

1. Application not handling SIGTERM signal (only responding to SIGKILL)
2. Grace period too short for application shutdown (database flushes, file closes, etc.)
3. Infinite loops or deadlocks in shutdown handlers
4. Long-running requests not cancelling during shutdown
5. init processes in containers masking signals
6. finalizers on objects blocking pod deletion
7. Sidecar containers (logging, monitoring) hanging during shutdown

How to fix "Pod grace period exceeded, force killing pod"

Advanced notes

Related errors

• Failed to connect to server: connection refused (HTTP/2)

  How to fix "HTTP/2 connection refused" error in Kubernetes

  HIGH

• missing request for cpu in container

  How to fix "missing request for cpu in container" in Kubernetes HPA

  HIGH

• error: invalid configuration

  How to fix "error: invalid configuration" in Kubernetes

  HIGH

Official resources & further reading

• 📚Kubernetes Official Documentation
• 📖
  Kubernetes API Reference(Official Docs)
• 📖
  Kubernetes Tasks: Run Applications(Official Docs)
• 💬
  Stack Overflow: Kubernetes grace period exceeded(Stack Overflow)
• 🐙
  Kubernetes GitHub Issues(GitHub)