action.skip

Upgrading using the Linux installer

When upgrading, it is important to remove any activation files from MSS associated with previous versions of Host Access for the Cloud. Leaving obsolete activation files in place may result in limited access to sessions.

What's required before upgrading?

  • Administrative privileges for the operating system.
  • The cluster will be unavailable for end users during the upgrade process. We recommend planning a maintenance window accordingly.
  • Each node in the cluster must be in a Ready state before you attempt to upgrade.

Upgrade steps

To upgrade your product, first ensure all nodes in the deployment are in a Ready state by running: cspctl status with elevated privileges. Then perform the following steps:

  1. From the Micro Focus download site, download the Linux installer script (install*.sh) for your product.

  2. Enable execute permissions for the installer:

    chmod 744 install*.sh

  3. On any any node in the cluster run: cspctl cluster cordon. The status of each node will change to Ready/SchedulingDisabled. This command can be safely skipped if it is not available, it is only available in more recent versions of the cspctl.

  4. On each node in the cluster, update one node at a time by repeating the following steps:

    a. Copy the installer to the node, run the Linux install script (.sh) with elevated privileges, (for example, sudo), to upgrade the product.

    b. After the upgrade is complete, the verification tool automatically runs.

    • If verification succeeds, the services will automatically start.
    • If verification fails, review the troubleshooting steps.

    c. After the CSP service starts, wait until the updated node shows a status of Ready/SchedulingDisabled.

    • Throughout the cluster upgrade process, it is normal to see warnings and errors in output of cspctl status. These will clear once the entire upgrade process is complete.
    • It can take up to 15 minutes for the node to become Ready.

    d. Move on to the next node.

  5. Once all nodes in the cluster have been updated and are ready, run: cspctl cluster scale on any node

  6. Wait for the cluster to return to a healthy state with all nodes showing Ready,SchedulingDisabled, or Ready if cordon was not used. This process can take up to 15 minutes.

  7. On any node run: cspctl cluster uncordon to allow all nodes to become schedulable again. This command can be safely skipped if it is not available.

  8. Once all nodes return to a Ready status, the cluster is ready for use.