NetBackup™ Upgrade Guide

Preinstall steps to upgrade to NetBackup 10.0 and complete the image metadata migration

1. Perform environment checks with the SORT tool.

   See Recommended SORT procedures for upgrades.

2. Download the Customer Registration Key for Veritas Usage Insights. More information about Veritas Usage Insights is available.

   See About Veritas Usage Insights.

   During install and upgrade to NetBackup 10.0, please allow the installer to copy the veritas_customer_registration_key.json file to its final destination. NetBackup can set the file permission and ownership correctly through this process. If you place the file onto your systems outside of the install or the upgrade process, the process may not work correctly.

3. (Conditional) On Linux, if the NetBackup database files reside on btrfs file system, move it to a supported file system (such as ext4 or xfs) before you upgrade. It is not supported to reside the NetBackup database on a btrfs file system. The database files reside on the primary server in the directories under /usr/openv/db.

   To move the NetBackup database files on Linux:

   • Perform a catalog backup.

   • Shut down all NetBackup daemons:

     /usr/openv/netbackup/bin/bp.kill_all

   • Start the SQL Anywhere daemon:

     /usr/openv/netbackup/bin/nbdbms_start_stop start

   • Move the existing data, index, and transaction log files:

     /usr/openv/db/bin/nbdb_move -data data_directory -index index_directory -tlog log_directory

     If you use mirrored transaction log, use the command shown:

     /usr/openv/db/bin/nbdb_move -data data_directory -index index_directory -tlog log_directory -mlog log_mirror_directory

   • Start all NetBackup daemons:

     /usr/openv/netbackup/bin/bp.start_all

   • Perform a catalog backup.

4. (Conditional) On Windows, be sure that all operating system and security updates are applied. More information is available.

   See Windows compiler and security requirements for NetBackup 9.1 and later installation and upgrade.

5. Perform any pre-upgrade tasks that you would normally do in regard to your NetBackup environment. For example:

   • Stop all customized or third-party scripts.

   • Perform any cluster-specific tasks.

   • Run a hot catalog backup.

   • Disable OpsCenter data collection for this primary server.

   • Disable all storage lifecycle policies (SLPs).

   • Deactivate all NetBackup policies.

   • Deactivate all disk staging storage units for all pre-NetBackup 7.5.x environments.

   • For VMware and Red Hat Virtualization (RHV) environments, reduce the size of the job database before upgrade. Following an upgrade, existing VMware and RHV jobs are assigned an asset namespace to enable access control at an asset level. This process may take some time. More information about this process is available:

     See Reduce the job database size before upgrade.

   • For clustered systems only, take the following NetBackup resources offline:

     • Windows Server Failover Clusters (WSFC): Take all of the NetBackup group resources offline except for the disk, the virtual name, and the virtual IP address. Refer to the Microsoft Cluster Administration documentation to determine how to take the NetBackup group resources offline through the cluster administrator interface.

     • Cluster Server (VCS) clusters: Take the NetBackup resource offline.

       Freeze the NetBackup group with the -persist option using the command shown:

       hagrp -freeze NetBackup_service_group -persistent

     Refer to the NetBackup Clustered Primary Server Administrator's Guide for the commands to take these resources offline.

6. (Conditional) If you plan to change your NetApp cluster to Vserver mode from node scope mode, create a detailed image report for each filer. You can generate this report with the bpimagelist command. The example that is shown is one possible option. Use whatever options are necessary for your environment.

   bpimagelist - client ndmp_host_name

7. Beginning with NetBackup 8.0, the NetBackup primary server includes a configured Tomcat web server to support critical backup operations. This web server operates under user account elements with limited privileges. These user account elements must be available on each primary server (or each node of a clustered primary server). More information is available:

   See NetBackup master server web server user and group creation.

   Note:

   It is recommended that you save the details of the user account that you use for the NetBackup Web Services. A primary server recovery requires the same NetBackup Web Services user account and credentials that were used when the NetBackup catalog was backed up.

   Note:

   If the NetBackup PBX is running in secure mode, please add the web service user as authorized user in PBX. More information about determining PBX mode and how to correctly add users is available.

   http://www.veritas.com/docs/000115774

8. (Conditional) If you have customized the settings for your Tomcat web server, determine if those settings persist across upgrades. More information is available:

   See Persistent Java Virtual Machine options.

9. Stop any applications on the system that interact with NetBackup. This step includes any databases or system components being backed up. Failure to stop these applications may result in unexpected behavior. Observed behavior includes aborted upgrades and application failures.

   For Oracle users, ensure that no backups are running. Stop the RMAN processes before you install NetBackup. If you use AIX, you should run /usr/bin/slibclean as the root user after you stop the RMAN processes.

   If you cannot stop your Oracle database, a procedure is available that may let you install NetBackup with the Oracle database active. More information on this topic is available.

   http://www.veritas.com/docs/TECH158276

10. Stop all NetBackup services.

    • On UNIX systems: /usr/openv/netbackup/bin/bp.kill_all

    • On Windows systems: install_path\NetBackup\bin\bpdown -f