Back to all scenarios
Scenario #355
Storage
Kubernetes v1.24, CSI

Pod Deleted but Volume Still Mounted on Node

Pod was force-deleted, but its volume wasn’t unmounted from the node, blocking future pod scheduling.

Find this helpful?
What Happened

Force deletion bypassed CSI driver cleanup. Mount lingered and failed future pod volume attach.

Diagnosis Steps
  • 1kubectl describe node showed volume still attached.
  • 2lsblk confirmed mount on node.
  • 3Logs showed attach errors.
Root Cause

Orphaned mount due to force deletion.

Fix/Workaround
• Manually unmounted the volume on node.
• Drained and rebooted the node.
Lessons Learned

Forced pod deletions should be last resort.

How to Avoid
  • 1Set up automated orphaned mount detection scripts.
  • 2Use graceful deletion with finalizer handling.
Previous ScenarioNext Scenario