> Software > Manage > Storage > HyperScale > HyperScale X > HyperScale X Upgrade

Upgrading the HyperScale X Version on Existing Nodes

You can upgrade HyperScale X 2.x clusters to HyperScale X 3.x clusters.

Review the following considerations:

  • The upgrade process is performed sequentially, and may require up to 90 minutes per node to complete. For example, a three-node cluster may require up to 4.5 hours to complete. For customers with systems configured with BIOS boot, additional time may be required.

  • The upgrade process will prompt to update the password for both the root and cvbackupadmin users.

  • The upgrade process automatically creates the cvbackupadmin user, if the user is not already present before the upgrade.

  • The upgrade process does not impact the drives that host Deduplication database (DDB) and Index Cache. The drives hosting backup data are also not modified during the upgrade.

  • Once the upgrade is complete, ransomware protection, restricted shell, and firewall are enabled by default, even if the storage pool did not have these settings enabled previously.

  • After the upgrade starts, it is not possible to rollback to the previous version.

  • To prevent a significant increase in the deduplication database and storage, avoid performing full backups and synthetic full operations during the upgrade process.

Before You Begin

  • If your hardware supports booting from NVMe drives, ensure that the /ws/ddb/iso path is located on a bootable NVMe drive.

    Note: NVMe drives containing the /ws/ddb/iso path appear in the boot order menu in the BIOS.

    Before starting the migration, update the BIOS boot settings to place the bootable NVMe drive above CVLTPBBA and at the top of the boot order.

  • The Managed Service Provider (MSP) administrator or the tenant administrator performing the installation must obtain the authorization code.

    For instructions on activating the authorization code, see User Administration and Security.

  • If a tape library is attached to the MediaAgent, complete the following steps:

    1. Do not use the tape library for any operations like backup, restore or auxiliary copy.

      • Verify that all tape library drives are unloaded.

      • Confirm the library is in an "Idle Ready" state.

    2. Remove lin_tape Driver (if applicable)

      • If using the lin_tape driver, remove it using the following command:

        rpm -e [lin_tape_version].rpm

        Example:

        rpm -e lin_tape-1.60.0-1.x86_64.rpm

      • If a lin_tape.conf file exists under /etc/modprobe.d, remove it.

Procedure

To upgrade the HyperScale X version, complete the following steps:

  1. Using the system console (for example, using KVM, not SSH), log on to the node in which the remote cache is configured, and then create and navigate to the /ws/ddb/iso folder.

    mkdir /ws/ddb/iso
cd /ws/ddb/iso

  2. Copy the latest HyperScale X image from Commvault store to the /ws/ddb/iso folder on the Hyperscale X remote software cache node.

  3. To verify if the existing cluster meets the upgrade criteria, run the following command:

    cd /opt/commvault/MediaAgent
./cvmanager.py -t Validate_Migrate_Cluster

  4. The validation process will prompt for the following inputs:

    • For SaaS environment:

      [root@ehsxn1 ~]# cd /opt/commvault/MediaAgent
[root@ehsxn1 MediaAgent]# ./cvmanager.py -t Validate_Migrate_Cluster
        :      Initializing HyperScale Task Manager.
        :      Commvault HyperScale Task Manager Initialized.
        :      No incomplete task data found for: Validate_Migrate_Cluster
Are you registering with SaaS (y/n): y
Enter the SaaS backup gateway for registration: 
Enter the SaaS backup gateway port for registration [443]: (443)
Enter the auth code for registration:

    • For Software environment:

      [root@ehsxn1 MediaAgent]# ./cvmanager.py -t Validate_Migrate_Cluster
        :      Initializing HyperScale Task Manager.
        :      Commvault HyperScale Task Manager Initialized.
        :      No incomplete task data found for: Validate_Migrate_Cluster
Are you registering with SaaS (y/n): n
CommServe fully qualified hostname:
Enter the auth code for registration:

    You must proceed with the upgrade only after the validation process is complete. If the validation fails, you must fix the issue and then rerun the validation command.

  5. Initiate the upgrade by executing the upgrade command.

    cd /opt/commvault/MediaAgent
./cvmanager.py -t Migrate_Cluster

  6. The migration script will prompt for the following inputs:

    • For SaaS environment:

      [root@ehsxn1 MediaAgent]# ./cvmanager.py -t Migrate_Cluster
        :      Initializing HyperScale Task Manager.
        :      Commvault HyperScale Task Manager Initialized.
        :      No incomplete task data found for: Migrate_Cluster
Root password of existing cluster:
[Re-Enter to Confirm] - Root password of existing cluster:
Restricted shell user (cvbackupadmin) password:
[Re-Enter to Confirm] - Restricted shell user (cvbackupadmin) password:
Are you registering with SaaS (y/n): y
Enter the SaaS backup gateway for registration: 
Enter the SaaS backup gateway port for registration [443]: (443)
Enter the auth code for registration:

    • For Software environment:

      [root@ehsxn1 MediaAgent]# ./cvmanager.py -t Migrate_Cluster
        :      Initializing HyperScale Task Manager.
        :      Commvault HyperScale Task Manager Initialized.
Root password of existing cluster:
[Re-Enter to Confirm] - Root password of existing cluster:
Restricted shell user (cvbackupadmin) password:
[Re-Enter to Confirm] - Restricted shell user (cvbackupadmin) password:
Are you registering with SaaS (y/n): n
CommServe fully qualified hostname: 
Enter the auth code for registration:

  7. The migration process migrates and reboots each node in the cluster from RHEL7 to Rocky8 sequentially. This process may take several hours to complete.

    Caution

    Do not attempt to manually upgrade any node in the cluster while migration is in progress.

  8. To verify whether the node was migrated successfully, on the system console, run the following command and make sure the output is displayed as follows:

    # commvault reg | grep MigrationCompleted

    Output:
    sRHELToRockyMigrationCompleted yes

  9. To resume a upgrade job automatically when there are incomplete tasks, start the upgrade using the following command:

    ./cvmanager.py -t Migrate_Cluster --resume
    When you run the upgrade without the resume parameter, the upgrade job detects the existing migrate_cluster task and offers options to resume the task.

What to Do Next

  1. Synchronize the remote cache after a successful migration. For instructions, see Synchronizing, Disabling, or Deleting an Existing Software Cache.

  2. If a tape library is attached to the MediaAgent, complete the following steps:

    1. From the Command Center navigation pane, go to Manage > Storage > Tape..

      The Tape page appears.

    2. In the upper-right corner of the page, click Add to add a tape storage.

      The Add tape dialog box appears.

    3. From the MediaAgent list, select the MediaAgent to which the tape library is attached.

    4. Click Scan hardware.

      Tail the cvd.log file on the MediaAgent or Node to monitor progress. The output of CVMACScsi::DetectDevices() will display the access paths.

      Once the scan is complete, close the Add tape dialog box.

    5. From the Command Center navigation pane, go to Manage > Storage > Tape.

      The Tape page appears.

    6. For the tape_storage, click the Actions button, and then click Reset library.

    7. Click Yes in the Confirm reset library prompt.

      The reset is complete when CVMA.log shows details on slots, drives, and IE ports, concluding with: WORKER [ ] libHandlerError = 0, bStatus = 1

×

Loading...