Chapter 1. Introduction to scaling storage

Red Hat OpenShift Data Foundation is highly scalable storage system. OpenShift Data Foundation allows you to scale by adding the disks in the multiple of three, or three or any number depending upon the deployment type.

  • For internal (dynamic provisioning) deployment mode, you can increase the capacity by adding 3 disks at a time.
  • For internal-attached (Local Storage Operator based) mode, you can deploy with less than 3 failure domains.

With flexible scale deployment enabled, you can scale up by adding any number of disks. For deployment with 3 failure domains, you will be able to scale up by adding disks in the multiple of 3.

For scaling your storage in external mode, see Red Hat Ceph Storage documentation.


You can use a maximum of nine storage devices per node. The high number of storage devices will lead to a higher recovery time during the loss of a node. This recommendation ensures that nodes stay below the cloud provider dynamic storage device attachment limits, and limits the recovery time after node failure with local storage devices.

While scaling, you must ensure that there are enough CPU and Memory resources as per scaling requirement.

Supported storage classes by default

  • gp2-csi on AWS
  • thin on VMware
  • ovirt-csi-sc on Red Hat Virtualization
  • managed_premium on Microsoft Azure

1.1. Supported Deployments for Red Hat OpenShift Data Foundation

  • User-provisioned infrastructure:

    • Amazon Web Services (AWS)
    • VMware
    • Bare metal
    • IBM Power
    • IBM Z or LinuxONE
  • Installer-provisioned infrastructure:

    • Amazon Web Services (AWS)
    • Microsoft Azure
    • Red Hat Virtualization
    • VMware