Use this procedure to upgrade Cluster Control Manager. The upgrade takes about 15 to 20 minutes to complete. When the upgrade completes, the upgrade utility exits and the Cluster Control Manager automatically reboots.
Before you begin
Ensure that file integrity validation is disabled. You can check this by running clusterFileIntegrity. To disable file integrity validation, run clusterFileIntegrity disable.
You can re-enable file integrity validation after the entire solution upgrade process is complete.
Follow the procedures in your solution documentation to take a full backup of the solution service data, export metric data, and logs (if needed).
Run ccm report to back up Elasticsearch data.
Run ccm release common-services exportPrometheusSnapshot to back up solution metric (Prometheus) data.
Using a utility, such as WinSCP, transfer the updated solution spreadsheet and upgrade-config.yaml file, if one is provided, to Cluster Control Manager.
Unless instructed by your solution documentation, do not update the values on the cluster-config tab of the spreadsheet.
If utilizing the Common Services High Availability Audit, verify that the vCenter credentials provided during the initial cluster installation are not expired. If these credentials are expired or close to expiring, update the credentials in vCenter and then update the credentials on Cluster Control Manager using the ccm infra update-vcenter-creds command.
From the vCenter that is managing the solution cluster, take a virtual machine snapshot of Cluster Control Manager.
Note:
VMware VM snapshots are supported for cluster nodes only when the nodes are powered off. Keep snapshots for a maximum of 72 hours. Over longer periods, snapshot files increase in size and can degrade performance of the virtual machine and the ESXi host.
(Optional): Prestage the CCM upgrade artifact.
Prestaging artifacts is non-service impacting and can be completed outside of a maintenance window.
Procedure
Log in to Cluster Control Manager using the customer account login.
Create the upgrade-config.yaml file.
Note:
If an upgrade configuration YAML file has been provided by your solution, use that file and skip this step.
Enter the vi /home/<customer account directory>/upgrade-config.yaml command.
Enter the version string from your solution documentation.
For information about the GA version number to use, see your solution release notes.
Save and exit the file.
From the directory on Cluster Control Manager that contains the upgrade-config.yaml file, enter the screen command.
The screen utility enables the upgrade to run in the background.
To begin the upgrade, run the ccm upgrade system upgrade-config.yaml command.
When prompted, enter your Avaya SSO credentials.
Accept the EULA.
You must accept the EULA for the upgrade to continue.
A snapshot warning displays indicating not to proceed with the upgrade if you have not taken a snapshot. A prompt also indicates that the system will reboot after the upgrade is applied.
At the Do you wish to continue? prompt, enter y to confirm that you have taken a snapshot and to acknowledge the reboot.
Important:
If a snapshot has not been taken, type n and return to the Before you Begin section to confirm that all prerequisites have been completed before continuing.
When prompted to take a backup after the CCM upgrade, enter y or n.
If you enter y, a backup runs after the upgrade is complete. The backup takes approximately 5 to 10 minutes to complete. If the archive destination is set to Local, the backup file is located at /var/avaya/artifactCache/ccmClusterBackup. If the archive destination is set to Remote, the ccmClusterBackup folder, which contains the backup file, is located in the base directory you specified.
Optional Monitor upgrade progress on the screen session or by viewing /var/log/avaya/ccm/upgrade.log using tail -f command.
Optional If the upgrade fails with an error, resolve the issue causing the failure and restart the upgrade by reverting to your snapshot and running the ccm upgrade system upgrade-config.yaml command.
Contact Avaya support personnel if the problem persists.
Wait for the cluster node power-on process to complete.
Wait approximately 30 minutes, then run ccm smoke-test on Cluster Control Manager and confirm that all tests pass.If ccm smoke-test fails, run it again in 10 minutes. If the test continuously fails for over an hour, contact your technical support representative.
