Skip to content

Datastores

Extra datastores can be useful if; for example, the Cluster has multiple "tiers" of storage; such as, NVMe, SSD, and HDD.

Warning

Real device classes should be used; however, for this example, we will only create named pools.

Alternatively, additional Datastores may be added if different resiliency strategies are required in the same Cluster. (e.g. Triple replication and 8+3 erasure coding.)

  1. SSH into the HyperCloud Dashboard.
  2. Create a pool.
    • ceph osd pool create <pool_name>
      • For example: ceph osd pool create nvme, where nvme = <pool_name>

  3. Login to the HyperCloud GUI.

    • Go to Storage -> Datastores.
    • Click the green dropdown button and select 'Create'.
    • Input logical designation for Name.
    • Select Ceph from the Storage backend dropdown list.
    • Add hypercloud-dashboard to the Host bridge list.
    • Input the name of the pool <pool_name> you just created in the HyperCloud Manager into the Ceph pool to store Images field.
    • Add hypercloud-storage-1 to Ceph host.

    Create Datastore Image

  4. Click green Create button.

  5. The new Datastore will now be available in the list.

    DataStore List Image

The easiest way to make use of this Datastore is to download the 'Empty disk' Image from the marketplace, give it the Name 'Empty NVMe Disk', and select the relevant Datastore.

Download Empty Disk App Image

This empty disk can now be added to VMs at instantiation or via a Template.