Unhealthy disk are seen on the Infrastructure page after you delete a node from the cluster
If an add node operation fails and you delete the new node from the cluster, stale disk entries for the deleted node are shown in the UI. The page shows unhealthy disks for the deleted node. After the add node operation fails, the recovery tasks run partial discovery, which does not remove these entries from the UI. (IA-46378)
Workaround:
Run full discovery by navigating to or wait for the full discovery to run automatically as scheduled. The status is updated automatically in the UI after the full discovery is completed.