Upgrading GUI high availability nodes

In a GUI high availability (HA) environment, upgrade all GUI nodes before you upgrade the cluster nodes by using the essrun CLI. Upgrade the GUI nodes one at a time so that at least one GUI node is available during the upgrade process.

Make sure that the upgrade packages are available on each GUI node and identify the master and other two GUI nodes.

A cluster with GUI HA can have three IBM Storage Scale GUI nodes. One of these nodes is called mater GUI node and the other two are called primary and secondary nodes.
Master GUI node
This node is automatically elected to complete cluster-wide tasks that must run only once in the cluster. These tasks include snapshot schedules, email notifications, and SNMP notifications. If the current master GUI node becomes unavailable, another GUI node is automatically elected as the master GUI node.
Primary and secondary nodes
One of these nodes can become a master node when the current master GUI node fails.
  1. Find the GUI nodes.
    /usr/lpp/mmfs/gui/cli/lsnode
    A sample output is as follows:
    
    Hostname             IP          Description     Role               Product version Connection status GPFS status Last updated
    gpfsgui-11.novalocal 10.0.100.12 Master GUI Node management,storage 7.0.1.0         HEALTHY           HEALTHY     7/10/17 10:19 AM
    gpfsgui-12.novalocal 10.0.100.13                 storage,ces        7.0.1.0         HEALTHY           HEALTHY     7/10/17 10:19 AM
    gpfsgui-13.novalocal 10.0.100.14                 storage,ces        7.0.1.0         HEALTHY           HEALTHY     7/10/17 10:19 AM
    where:
    gpfsgui-11.novalocal
    Is a master GUI node that fails over automatically.
    gpfsgui-12.novalocal
    Is the primary GUI node in other nodes.
    gpfsgui-13.novalocal
    Is the secondary GUI node in the other nodes.
  2. Stop the GUI service on the secondary GUI node.
    systemctl stop gpfsgui-13.novalocal

    Verify that the GUI service is stopped.

    systemctl status gpfsgui-13.novalocal
  3. Stop the GUI service on the primary GUI node.
    systemctl stop gpfsgui-12.novalocal

    Verify that the GUI service is stopped.

    systemctl status gpfsgui-12.novalocal
  4. Upgrade the primary GUI node.

    Install the required IBM Storage Scale GUI packages and complete any release-specific postinstallation steps.

    rpm -Uvh gpfsgui-12.novalocal*.rpm
  5. Start the GUI service on the primary GUI node.
    systemctl start gpfsgui-12.novalocal

    Verify that the GUI service is running and that all GUI nodes are healthy.

    systemctl status gpfsgui-12.novalocal
    
    /usr/lpp/mmfs/gui/cli/lsnode
  6. Upgrade the secondary GUI node.

    Install the required IBM Storage Scale GUI packages and complete any release-specific postinstallation steps.

    rpm -Uvh gpfsgui-13.novalocal*.rpm
  7. Start the GUI service on the secondary GUI node.
    systemctl start gpfsgui-13.novalocal

    Verify that the GUI service is running and the node is healthy.

    systemctl status gpfsgui-13.novalocal
    
    /usr/lpp/mmfs/gui/cli/lsnode
  8. After all GUI nodes are upgraded and verified, start the cluster upgrade by using the essrun command.
  9. After an IBM Storage Scale System upgrade is complete, start the GUI.