Scale out protection for virtual machines
Updated: May 13, 2016
Applies To: System Center 2012 SP1 - Data Protection Manager, System Center 2012 R2 Data Protection Manager
In System Center 2012 Service Pack 1 (SP1) - Data Protection Manager (DPM), you can scale out protection of Hyper-V clusters. The new scale-out feature removes the limit of a one-to-one relationship between a Hyper-V cluster and a DPM server, and the DPM protection agent that runs on a virtual machine can attach itself to multiple DPM servers. Therefore you can add the virtual machine to a protection group on any of the recognized DPM
This feature is only available for virtual machines that are hosted on servers that are running Windows Server 2012.
The scale-out feature supports protection for remote SMB 3.0-complaint shares for a virtual machine.
The DPM server must be a clean deployment of DPM in System Center 2012 SP1.
DPM chaining is not supported.
All DPM servers and the protected computers must be located in the same domain.
The scale-out feature does not support an IPv4 or IPv6 address for a storage location file.
To deploy scale-out protection, you must have a minimum of two DPM servers D1 and D2, which are visible to all virtual machines that are hosted on nodes N1, N2, N3, N4, and N5. When you create protection groups on D1 or D2, you can add any of the virtual machines from V1 to V10 for protection.
Deploy DPM—Verify that DPM is installed and deployed correctly. If you haven’t see:
Set up storage—Check that you have storage set up. Read more about your options in:
Learn about short-term storage to disk and storage pools in Plan for disk backups
For storage to Azure with Azure Backup, see Plan for Azure backups
For long-term storage to tape, Plan for tape-based backups
Use the following capacity sizing for virtual machine protection.
Average virtual machine size
100 gigabytes (GB)
Number of virtual machines per DPM server
Total size of 800 virtual machines
80 terabytes (TB)
Required space for backup storage
80 terabytes (TB)
As an example, if your fully scaled Hyper-V cluster contains 1600 virtual machines, you’ll need two DPM servers, and double the provisioning that is summarized in the table. The total storage requirement for 1600 virtual machines is greater than 160 TB. Read about optimizing space in Optimize virtual machine protection.
Set up the DPM protection agent—The agent needs to be installed on the Hyper-V server, or on each server in the Hyper-V cluster. Read Plan for protection agent deployment, and then Set up the protection agent.
Make multiple servers visible—Use the SetDPMServer command with the Add parameter on the protected virtual machine to make multiple DPM servers visible to the protected virtual machine, as follows: Setdpmserver –add –dpmservername <name of second DPM server>. For more information about this command, see Using SetDPMServer [DPM2012_Web].
If you don’t use the –Add parameter, the previous DPM server is overwritten. Ensure that all servers that are running Hyper-V and virtual machines are discovered by the Virtual Machine Manager (VMM) server before you begin to create protection groups.
Set up a protection group—In the Select Group Members page of the Create New Protection Group wizard you can select the Hyper-V host servers on which the virtual machines you want to back up are located. Keep virtual machines to which you want to apply the same protection policy in the same protection group. If you have existing protection groups, after you attach all the DPM servers, add the virtual machine to a protection group on any of the DPM servers that the virtual machine recognizes. After you add the virtual machine to a group, the virtual machine is always backed up on that DPM server. If you want the virtual machine to be backed up on a different server, you must stop protection for it and add it to a protection group on the new server. This flexibility means that virtual machines from a single node can be protected by different DPM servers. Read about protection groups in:
Then follow the instructions in Create and manage protection groups.
Configure new nodes for protection— If you want to add a new node to the cluster, you must install the DPM protection agent on that node, and then configure it for protection, as follows:
Install the DPM protection agent on the server that is running Hyper-V.
Attach the DPM protection agent to all DPM servers in the network.
Run Setdmpserver on all nodes in the cluster.
Add a new DPM server— If you want to add a new DPM server to the scaled-out deployment, it must be configured with all the nodes in the cluster to prepare it for protection. To do this attach a new node for protection, and run Setdpmserver on all nodes in the cluster.
After you create the protection group initial replication of the data occurs. Backup then takes place in line with the protection group settings. If you need to recover backed up data see Recover backed up data [DPM2012_Web]. Note the following behavior for restoring virtual machines.
Virtual machine recovery
Recover to original location
The original VHD is deleted. DPM will recover the VHD and other configuration files on the original location using Hyper-V VSS writer. At the end of the recovery process, virtual machines will still be highly available.
The resource group must be present for recover. If it isn’t available recover to an alternate location and then make the virtual machine highly available.
Recover to alternate location
DPM supports alternate location recovery (ALR), which provides a seamless recovery of a protected Hyper-V virtual machine to a different Hyper-V host, independent of processor architecture. Hyper-V virtual machines that are recovered to a cluster node will not be highly available.
Item-level recovery (ILR)
DPM supports item-level recovery (ILR), which allows you to do item-level recovery of files, folders, volumes, and virtual hard disks (VHDs) from a host-level backup of Hyper-V virtual machines to a network share or a volume on a DPM protected server. The DPM protection agent doesn’t have to be installed inside the guest to perform item-level recovery.