For OCS clusters with service capacity of 4TiB a custom storage class should be generated with cachingmode:none. Otherwise, the deviceset pods with remain in pending with an error message concerning the acces mode. A note should be added here to address that issue. Reported by: rhn-gps-swilson https://access.redhat.com/documentation/en-us/red_hat_openshift_container_storage/4.5/html/deploying_and_managing_openshift_container_storage_using_microsoft_azure/deploying-openshift-container-storage-on-microsoft-azure_azure#annotations:ae1308ac-0a62-411f-9b47-bd4f43a6f143
Example storage class definition: ``` provisioner: kubernetes.io/azure-disk kind: StorageClass apiVersion: storage.k8s.io/v1 metadata: name: managed-premium-over4ti parameters: kind: managed storageaccounttype: Premium_LRS cachingmode: None reclaimPolicy: Delete allowVolumeExpansion: true volumeBindingMode: WaitForFirstConsumer ```