23_check_storage_health.yml - Usage Guide

Read-Only

Purpose

Read-only storage report on each host:

  1. df -hT — real filesystems
  2. df -i — inode usage
  3. containerd/crictl / docker disk usage + total image size
  4. kubectl if present: PVC / PV / StorageClass + unbound PVCs

Host scope

hosts: all — disk/inode is local to each node. PV/PVC data is filled on the node with kubectl access (usually the control-plane); if kubectl is missing on a worker, that section is skipped.

How to run 

ansible-playbook -i inventories/cagatayuresincom/hosts.ini playbooks/23_check_storage_health.yml

ansible-playbook -i inventories/musteri_a/hosts.ini playbooks/23_check_storage_health.yml --limit workers

Interpretation

Finding Action
High disk Use% Log / image accumulation; 17–18
High IUse% Many small files / layers
PVC Pending StorageClass / provisioner / quota
PV Released/Failed Manual cleanup / reclaim policy

Notes

  • Script: playbooks/files/storage_health_check.py
  • become: true (du / runtime directories)
  • Does not change the cluster.