Adding expansion racks

You can expand your IBM Storage Fusion HCI System with an additional rack.

Before you begin

The IBM Service Support Representative must have completed the setup stage installation for the expansion rack.

Procedure

  1. From the IBM Storage Fusion HCI System menu, click Infrastructure > Nodes.
  2. Go to Actions and click Add rack.
    For Compute-storage nodes, the nodes on the auxiliary rack must have a minimum of two NVMe or same number of NVMes as that of the base rack.
    The Add rack wizard page is displayed.
  3. In the Getting started wizard page, go through the details and click Next to initiate the discovery process.
  4. In the Discovery wizard page, wait for the rack discovery to complete.
    In this page, IBM Storage Fusion HCI System enables expansion of rack, connects to the spine switch, connects to the provisioned node, and discovers rack data.
    Errors might occur during the discovery step because of the following reasons:
    • Unable to connect to the spine switch
    • Unable to establish connection to the provisioned node on the new rack. To resolve this issue, revisit your DNS configuration and retry.
    • Unsupported rack configuration. For example, the new rack contains six storage nodes, but eleven storage nodes are required as the existing storage configuration is using 8 + 3p erasure coding. Contact your IBM Support to identify the supported configuration.

    There are Retry and Cancel options on this page. If the discovery fails, it can be retried by clicking retry and if you wish to cancel, click Cancel.

  5. Click Next.
  6. In the Validation wizard page, check whether all the nodes are validated properly.
    The IBM Storage Fusion HCI System validates the configuration of each node as a preparatory step so as to join the cluster. A minimum of six nodes are required to add the rack successfully.

    This page continuously shows the number of nodes that are successfully validated. If the validation does not complete within the stipulated time, the validation operation fails and message corresponding to that gets displayed.

    If issues exist, then check your connections or fix problematic nodes and retry. You cannot proceed with errors, so click Save & Quit. Click Collect logs to download and view logs.
    In this page, if rack addition fails with the following error message, then contact IBM Support: 
    Rack import timedout

    The Retry on this page refreshes the node validation status.

  7. If the session is lost during the add rack operation, do the following steps:
    1. Log in to IBM Storage Fusion HCI System user interface.
    2. Go to Infrastructure > Nodes.

      A Scale out operation not completed notification is displayed on the page.

    3. Click Resume process link in the Scale out operation not completed notification.

      The Resume process allows you to continue the add rack operation from where you left.

      For any errors in rack addition due to timedout issues, go to the Infrastructure > Nodes > Discovered nodes tab page to know more details.

  8. Click Next.
    The optional Node selection wizard page is displayed.
  9. Optionally, in the Node selection wizard page, you can open the Advanced options and edit node assignments.
    By default, IBM Storage Fusion HCI System automatically configures all nodes in the rack. The Advanced options shows a table where you can deselect nodes. If you do not deselect, then the nodes get added automatically to the OpenShift® cluster. Your deselection depends on the amount of nodes required for your storage building block. For example, you have twenty nodes and six nodes are required for storage building block. The table displays only the remaining fourteen nodes. Also, a summary section is displayed with the following details:
    • Nodes
    • CPUs
    • Memory (GB)
    • Usable capacity (TiB)
    The values of the summary section changes dynamically based on your selection and deselection.
  10. Click Next.
    The Finish wizard page is displayed. The IBM Storage Fusion HCI System adds these nodes to the OpenShift cluster.
  11. In Finish wizard page, click Finish to close the Add rack wizard.

    After the add rack operation is complete, the user interface shows auxiliary rack related nodes in the Discovered node tab. To add more racks, click Actions > Add rack again and follow the steps mentioned in this procedure.

  12. Select nodes and click Add to cluster.
    The IBM Storage Fusion HCI System starts to add nodes to the OpenShift cluster. To know the progress of the OpenShift cluster addition, go to Configured nodes tab. If error occurs at this stage, a link is available to view nodes and errors in the error notification. For example, OpenShift experienced a timeout error. You can also click Retry to initiate the process. After the nodes are successfully added to OpenShift, the State changes to Ready for storage.
  13. When all the nodes are ready for storage configuration, click Add nodes to the storage cluster link in the notification. If you are not in the Nodes page, then a global notification is displayed and you can click the Configure storage link and return to the Nodes page.
    The State of the node changes to Installing storage software. After all the nodes are successfully added to storage, the following success notification is displayed and the State of the node changes to Running: 
    Rack expansion complete
All nodes are available for scheduling workloads