Node firmware

IBM Fusion HCI continuously monitors the firmware installed on each node. If a node is found to be running a firmware version lower than the Fusion-certified level, the system automatically initiates a firmware upgrade. If the upgraded firmware requires a reboot to take effect, the node is marked as Reboot Pending. In such cases, you must plan to manually trigger the reboot to apply the changes.

About the task

This automated firmware management applies to the following node types:
  • Nodes in the base OpenShift® Container Platform cluster
  • Nodes in Hosted Control Plane clusters
  • Service nodes
  • Nodes not part of any cluster
For service nodes and unclustered nodes, IBM Fusion HCI automatically handles the reboot process.
You can verify the status of the firmware in the Infrastructure > Nodes page. In the Firmware level column, you can find the firmware status that is associated with the release. In the following table, you can see some of the possible firmware statuses:
Status Description
Up to date It indicates that the node firmware matches the Fusion-certified level.
Update required Node is running outdated firmware; upgrade is orchestrated automatically.
Update failed Firmware upgrade failed; contact IBM support for assistance.
Reboot required Firmware has been staged; a reboot is needed for changes to take effect.

Procedure

  1. Go to Infrastructure > Nodes in the IBM Fusion HCI UI.
  2. Identify nodes marked as Reboot Required.
  3. Do the following steps to reboot the nodes:

    For the base OpenShift Container Platform nodes:

    1. Go to the Configured tab.
    2. Select the nodes and click Reboot.
    For Hosted Control Plane nodes:
    1. Go to the Hosted inventory tab.
    2. Select the nodes and click Reboot.
    Note: Do not trigger the Add Node to Cluster action for nodes listed under the Discovered tab if their firmware upgrade is still in progress.

    IBM Fusion HCI performs pre-reboot checks to ensure it is safe to proceed. If conditions are met, nodes are rebooted in a rolling fashion to minimize impact on cluster operations.