Why does mounting Block Storage for Classic fail with a file system check error?

Classic infrastructure

When you view pod logs for pods that are using Block Storage for Classic, you see an error similar to the following example.

UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY.(i.e., without -a or -p 
options)

The disk is corrupted.

Run the following commands to patch the PV that the pod is mounting.

1. Get the details of the pod that is failing and make a note of the PV and PVC it is mounting.

   kubectl get pods
   

2. Patch the PV to add the "ibm.io/autofix-fsckErr":"true" annotation.

   kubectl patch pv PVNAME -p '{"metadata": {"annotations":{"ibm.io/autofix-
   fsckErr":"true"}}}'
   

3. Delete and re-create your app pod to retry mounting the PV. After deleting and re-creating the pod, Kubernetes retries the mount operation and the Block Storage for Classic driver pod tries to auto fix the issue based on the "ibm.io/autofix-fsckErr":"true" annotation.

4. After the pod is successfully deployed and running, run the following command to edit the PV.

   kubectl edit pv PVNAME
   

5. Remove the "ibm.io/autofix-fsckErr: true" annotation from the PV and save it.

6. If the issue persists, contact support. Open a support case. In the case details, be sure to include any relevant log files, error messages, or command outputs.