Migrating to an incremental-forever incremental backup strategy

Use this procedure to migrate existing backup schedules, policies, and data mover nodes for use in an incremental forever backup strategy.

1. Upgrade all vStorage Backup Servers protecting a single vCenter. Make sure that this upgrade is completed at the same time for all data mover nodes.
   
   • This upgrade requires installing IBM Spectrum Protect™ Data Mover version 6.4 (or later) on the vStorage Backup Server.
   • As a discrete task, you do not have to complete Step 2 or Step 3 immediately following Step 1. After upgrading the data mover nodes, you can continue to back up VMs in your existing environment. You can complete Step 2 and Step 3 when a more convenient opportunity becomes available.
   Tip: If your environment uses multiple vStorage Backup Servers, consider upgrading only one server. Then, verify that your server operates successfully before upgrading the remaining vStorage Backup Servers.
2. Update the backup policy and backup schedules to implement incremental-forever incremental backups:
   Complete the following backup policy tasks on the IBM Spectrum Protect server by issuing commands in the administrative command-line client (dsmadmc):
   1. Create a management class for the appropriate domain and policy set for your incremental-forever incremental backups. This example creates management class mgmt_ifincr28 for domain domain1 and policy set prodbackups. The management class name is used to describe an incremental-forever incremental backup strategy that retains 28 backup versions:

      define mgmtclass domain1 prodbackups mgmt_ifincr28 description="Retain 28 backup versions"

   2. Create a backup copy group for your incremental-forever incremental backups. This example creates a standard backup copy group for domain domain1, policy set prodbackups, and management class mgmt_ifincr28:

      define copygroup domain1 prodbackups mgmt_ifincr28 standard type=backup

      The standard type=backup entries are default values and are not required to be specified. They are included in this example to illustrate that the copy group name is STANDARD and that the type of copy group is backup (instead of archive).
   3. Update the backup copy group with the appropriate version, retention, and expiration settings:
      
      Remember: In Data Protection for VMware version 6.2 and 6.3, backup version, retention, and expiration is based on a backup-chain granularity level. This method means that even though both incremental-forever full and incremental-forever incremental backups are taken (as part of the 6.2 and 6.3 incremental-forever full backup strategy), version expiration counts only full backups. In Data Protection for VMware version 6.4 (or later), backup version, retention, and expiration is based on a single-backup granularity level. This method means that version expiration counts both incremental-forever full and incremental-forever incremental backups.
      
      
      

      The verexists parameter specifies the maximum number of VM backup versions to retain on the server. If an incremental-forever incremental backup operation causes the number to be exceeded, the server expires the oldest backup version that exists in server storage. This example specifies verexists=28. This value means that a maximum of 28 VM backup versions are retained on the server.

      
      
      

      The retextra parameter specifies the maximum number of days to retain a VM backup version, after that version becomes inactive. This example specifies retextra=nolimit. This value means that the maximum number of inactive VM backup versions are retained indefinitely. However, when verexists is specified, the nolimit value is superseded by the verexists value. As a result, in this example, a maximum of 28 inactive VM backup versions are retained on the server.

   Based on the settings described in this step, the backup copy group is updated as follows:

   update copygroup domain1 prodbackups mgmt_ifincr28 verexists=28 retextra=nolimit

   
   In this example, the existing Data Protection for VMware version 6.3 environment consists of the following hosts and schedules:
   • An ESX cluster (esxcluster) that contains two ESX hosts (esxhost1, esxhost2).
   • The bup_esxcluster_full schedule runs a weekly incremental-forever full backup of each ESX host with data mover node dm1.
   • The bup_esxcluster_incr schedule runs a daily incremental-forever incremental backup of each ESX host with data mover node dm2.
   Complete the following backup schedule tasks in the Data Protection for VMware vSphere GUI:
   1. Start the Data Protection for VMware vSphere GUI by clicking the icon in the Solutions and Applications window of the vSphere Client.
   2. In the Getting Started window, click the Backup tab to open the Managing backup schedules window.
   3. Locate the backup schedule (used for incremental-forever full or incremental backups) to update. In this procedure, the incremental-forever full bup_esxcluster_full schedule is used.
   4. Right-click the schedule and select Properties.
   5. Go to the Schedule page and specify Incremental from the Backup strategy drop-down list.
   6. Click OK to save your update.
   7. Locate the backup schedule used for incremental-forever incremental backups. Right-click the schedule and select Delete. Since the incremental-forever full bup_esxcluster_full schedule was updated to incremental-forever incremental, this incremental-forever incremental schedule is no longer needed.
3. Now that you have an incremental-forever incremental backup schedule, you can reduce the number of data mover nodes by consolidating them:
   
   This example consolidates two data mover nodes into one data mover node.
   1. On the vStorage Backup Server, open a command prompt and go to the directory where the options file for dm1 is located.
   2. Using a text editor (such as Notepad), update this file with the following options:
      1. Specify vmmaxparallel to control the number of VMs backed up at one time by dm1:
         

         vmmaxparallel=2

         The default value and minimum value are 1. The maximum value is 50.
         Tip: For every data mover node you remove, increase the vmmaxparallel value by 1.
         
         
         Alternatively, you can specify vmlimitperhost to control the number of VMs backed up at one time by dm1 from the same ESX host:
         

         vmlimitperhost=1

         This option is useful when wanting to prevent a host from being overloaded. The default value is 0 (no limit). The minimum value is 1. The maximum value is 50.
   3. Log on the IBM Spectrum Protect server. Use the administrative command-line client (dsmadmc) to specify the maximum number of simultaneous VM backup sessions that can connect with the server. For example:
      
      

      maxsessions=4

      The default value is 25. The minimum value is 2.
4. Verify that the updated data mover nodes are working properly:
   1. Start the Data Protection for VMware vSphere GUI by clicking the icon in the Solutions and Applications window of your vSphere Client.
   2. In the Getting Started window, click the Configuration tab to view the Configuration Status page.
   3. In the Configuration Status page, select the vCenter that is protected in Step 1. Click a data mover node to view its status information in the Status Details pane. When a node displays a warning or error, click that node and use the information in the Status Details pane to resolve the issue. Then, select the node and click Validate Selected Node to verify whether the issue is resolved. Click Refresh to retest all nodes.