IBM Tivoli Storage Manager, Version 7.1

Virtual machine backup techniques

You can use several backup techniques to back up virtual machines.

Use the following information to determine which virtual machine backup technique to use that best meets your needs and environment.

Windows Hyper-V backup

The client can back up virtual machines that are hosted by a Hyper-V server on Windows systems.

How it works
Full virtual machine backup processing backs up and restores snapshots of Hyper-V virtual machines by using the Microsoft Volume Shadow Copy Services (VSS) interface. You can restore a virtual machine by using the files that are created by VSS.
Advantages
Hyper-V backup processing has the following advantages:
  • This method can back up data without stopping the virtual machine or stopping any running applications within the virtual machine when the guest virtual machine is running a VSS-enabled Windows operating system.
  • This method can restore either individual virtual machines or a group of virtual machines that run on a Hyper-V server for disaster recovery purposes.
  • This method adds backup and restore functions for guest operating systems without the need for you to install a Tivoli® Storage Manager client on the guest virtual machine.
  • This method can be used for both disaster recovery and long-term data backup support.
Disadvantages
Hyper-V backup processing has the following disadvantages:
  • Backups are not granular.
  • You cannot run individual file restore operations from a full virtual machine backup.

VMware full virtual machine backup

Full virtual machine backup processing backs up a virtual machine from a VMware ESX or ESXi-based host.

How it works
Full virtual machine backup processing stores a backup copy of all virtual disk images and configuration information for a virtual machine.
Advantages
With full virtual machine backup processing, you get faster data movement than a file-level backup.
Disadvantages
Full virtual machine backup processing has the following disadvantages:
  • Backups are not granular.
  • Full virtual machine backup operations enable a complete restore of a virtual machine, but they take more time and more server space than a file-level or incremental backup.
  • You can restore individual files from a full virtual machine backup only with IBM® Tivoli Storage Manager for Virtual Environments.
  • This method is only available on Linux and Windows clients.

VMware incremental backup

You can run an incremental backup of a virtual machine from a VMware ESX or ESXi-based host. Incremental backup processing requires a Tivoli Storage Manager for Virtual Environments license.

How it works
An incremental backup of a virtual machine backs up all changes that occurred since the previous backup of the virtual machine, whether the backup was a full backup, or another incremental backup.
Advantages
Incremental backup processing backs up changes to virtual machines between full virtual machine backups.
Disadvantages
Incremental backup processing of a virtual machine has the following disadvantages:
  • The size of incremental backups can increase if you do not run a full backup regularly.
  • It is inefficient to restore data from incremental backups because the process must automatically complete the following tasks:
    • Restore the most recent full backup.
    • Restore each incremental backup up to the specified recovery point.
  • This method is available only on Linux and Windows clients.

VMware incremental-forever-full virtual machine backup

Incremental-forever-full virtual machine backup processing backs up all the used blocks on a VMware virtual machine's disks. To run this type of backup, you must have a license to use Tivoli Storage Manager for Virtual Environments V6.4 or later.

How it works
The following processes occur during incremental-forever-full virtual machine backup processing:
  • A full virtual machine backup is required only one time.
  • Data from incremental backups is combined with data from the full backup to create a synthetic full backup image. This type of full backup is called synthetic because it is created from the data that is stored on the server and not from reading the used blocks on the production disks.
  • Subsequent incremental-forever-full virtual machine backup operations read all used blocks and copy those blocks to the server.
  • Each incremental-forever-full virtual machine backup operation reads and copies all of the used blocks, whether the blocks are changed or not since the previous backup.
Advantages
Incremental-forever-full virtual machine backup processing has the following advantages:
  • Periodic full backups are no longer necessary.
  • During a restore operation, you can specify options for a point in time and date to recover data. The data is restored from the original full backup and all of the changed blocks that are associated with the data.
Disadvantages
Incremental-forever-full virtual machine backup processing has the following disadvantages:
  • If one or more of the progressive incremental backups is corrupted on the server, you might not be able to fully recover a virtual machine. To ensure that you can fully recover a virtual machine, periodically run a full virtual machine backup.
  • This method is available only on Linux and Windows clients.

VMware incremental-forever-incremental backup

Incremental-forever-incremental backup processing backs up only the disk blocks that have changed since the last backup. To run this type of backup, you must have a license to use Tivoli Storage Manager for Virtual Environments V6.4 or later.

How it works
The following processes occur during incremental-forever-incremental backup processing of a VMware virtual machine:
  • A full virtual machine backup is required only one time.
  • A full virtual machine backup operation copies all of the used disk blocks that are owned by a virtual machine to the Tivoli Storage Manager server.
  • After the initial full backup, all subsequent backup operations of the virtual machine are incremental-forever-incremental backups.
  • This method copies only the blocks that changed since the previous backup, regardless of the type of the previous backup.
  • The server uses a grouping technology that associates the changed blocks from the most recent backup with data that is already stored on the server from previous backups.
  • A new full backup is then effectively created each time changed blocks are copied to the server by an incremental-forever-incremental backup.
Advantages
Incremental-forever-incremental backup processing has the following advantages:
  • Periodic full virtual machine backups are no longer necessary.
  • This method reduces the amount of data that goes across the network.
  • This method reduces data growth because all incremental backups contain only the blocks that changed since the previous backup.
  • No comparison with the backup target is required since only changed blocks are identified.
  • Impact to the client system is minimized.
  • The length of the backup window is reduced.
  • Data restore operations are simplified.
  • This method optimizes data restore operations.
Disadvantages
Incremental-forever-incremental backup processing has the following disadvantages:
  • If one or more of the progressive incremental backups is corrupted on the server, you might not be able to fully recover a virtual machine. To ensure that you can fully recover a virtual machine, periodically run a full virtual machine backup.
  • It is available only on Linux and Windows clients.

VMware file-level backup on Windows

On Windows systems, you can use the backup-archive client to create file-level backups of VMware virtual machines.

How it works
The following processes occur during file-level backup processing of a virtual machine:
  • A VMware snapshot is taken of the virtual machine to be backed up.
  • The file systems of the virtual machine are remotely mapped to the vStorage backup server.
  • A file-level progressive incremental backup is run for all of the file systems.
  • The data is stored under the node name that matches the host name of the virtual machine.
  • The data for each virtual machine is stored in the node that is associated with the virtual machine.
  • The file system mappings are removed and the snapshot is removed.
When to use
Use file-level virtual machine backup processing if you want the ability to restore individual files from a virtual machine but you do not have a license for Tivoli Storage Manager for Virtual Environments.
Advantages
File-level virtual machine backup processing has the following advantages:
  • You can use include and exclude rules to identify the files to back up.
  • Files are backed up as individual files rather than as an image backup.
Disadvantages
File-level restores must be made from a backup-archive client that is installed directly on a virtual machine guest.

Parallel backups of VMware virtual machines

You can improve performance of virtual machine backups by running parallel backups of multiple virtual machines by using a single instance of the backup-archive client. This feature is only available in the Tivoli Storage Manager V6.4 or later client.

How it works
The following processes occur during parallel backup processing of virtual machines:
  • A single Tivoli Storage Manager data mover node can be used to concurrently back up multiple virtual machines.
  • When the backups are initiated, the client establishes parallel sessions to copy the data to the Tivoli Storage Manager server.
Advantages
Parallel virtual machine backup processing has the following advantages:
  • The backup window is reduced.
  • You can optimize the backups so that they do not adversely affect the ESX servers that are hosting the virtual machines.
Disadvantages
You must optimize the parallel backups. The number of virtual machines that you can back up in parallel depends on the following factors:
  • The processing power of the vStorage server that the Tivoli Storage Manager data mover node runs on.
  • The performance of I/O between the client and the Tivoli Storage Manager server.


Feedback