After installing your kURL cluster, you might encounter an issue where the openebs-localpv-provisioner pod keeps failing or restarting. Consequently, related Persistent Volume Claims (PVCs) remain in the Pending state without being bound to any volume.
Before taking any recovery or cluster reset actions, it is recommended to first collect relevant diagnostic information about the openebs-localpv-provisioner pod. This will help you identify the root cause and take appropriate steps.
1. Collect Logs from the openebs-localpv-provisioner Pod
Run the following commands to describe the pod and collect its logs:
kubectl describe pod <openebs-localpv-provisioner-pod>
kubectl logs <openebs-localpv-provisioner-pod>
Look for errors that may indicate issues with the underlying host disk or other system factors.
2. Check the Status of the init-pvc-xxxxx Pod
Use the following command to check the status of the initialization pod associated with the PVC:
kubectl describe pod init-pvc-xxxxx
If this pod is not marked as Completed, it indicates that the PVC binding process has failed. Inspect the pod logs for more details on the failure.
3. Verify File Permissions in the /var/openebs Directory
Ensure that the permissions on the target directory are correct:
ls -ld /var/openebs
A correct example of the directory permissions should look like:
drwxr-xr-x
4. Collect a Host Support Bundle (if necessary)
To assist further troubleshooting, collect a host support bundle and share it with the Replicated support team for detailed analysis.