Share via


Plan for disk backups

 

Updated: September 19, 2017

Applies To: System Center 2012 SP1 - Data Protection Manager, System Center 2012 - Data Protection Manager, System Center 2012 R2 Data Protection Manager

Data Protection Manager (DPM) provides short-term backup to disk by saving data to the DPM storage pool. The storage pool is the set of disks on which the DPM server stores the replicas and recovery points for the protected data. Before you can start protecting data by using disk storage, you must add at least one disk to the storage pool.

The storage pool may consist of any of the following resources:

  • Direct attached storage (DAS)

  • Fiber Channel storage area network (SAN)

  • iSCSI storage device or SAN

Notes:

  • The DPM server must have at least two disks installed: one dedicated to the operating system and DPM installation files; and one dedicated to the storage pool. In the context of DPM, “disk” is defined as any disk device that’s manifested as a disk in the Windows Disk Management tool. DPM does not add any disk containing startup files, system files, or any component of the DPM installation to the storage pool.

  • Disks added to the storage pool should not have any partitions. To prepare disks for data protection, DPM will convert the disks to Dynamic.

  • The storage pool supports most disk types, including these:

    • Integrated Drive Electronics (IDE)

    • Serial Advanced Technology Attachment (SATA)

    • SCSI

  • Both the master boot record (MBR) and GUID partition table (GPT) partition styles are supported. We strongly recommend that you use GPT disks for the DPM storage pool as GPT disks provide additional flexibility of extending the disk size.

  • If you use a SAN for the storage pool, we recommend that you create a separate zone for the disk and tape used on DPM. Do not mix the devices in a single zone.

  • DPM does not support USB/1394 disks in the DPM storage pool. However, Virtual Tape Library (VTL) technology can be used to overcome this limitation. Use of a VTL is supported.

  • You cannot use storage spaces for the DPM disk storage pool.

  • Support for deduplicated volumes depends on the DPM operating system. For more information, see Storage issues.

  • DPM running as a Hyper-V virtual machine can store backup data to VHDs/ VHDXs in shared folders on a Windows file server with data deduplication enabled. For more information, see Deduplicate DPM storage.

  • Some original equipment manufacturers (OEMs) include a diagnostic partition that is installed from media that they provide. The diagnostic partition might also be named the OEM partition or the EISA partition. EISA partitions must be removed from disks before you can add the disk to the DPM storage pool.

  • You can also substitute custom volumes that you define in Disk Management for volumes in the storage pool.

  • The disks that are added to the DPM storage pool must all have the same sector size. You cannot mix 512-byte sector size disks with 4K sector size disks. This restriction stems from the way in which volume shadow copy (VSS) snapshots are taken and maintained.

Planning the storage pool involves the following:

  • Calculate capacity requirements

  • Plan the disk configuration

  • Define custom volumes

Calculate capacity requirements

Capacity requirements for the DPM storage pool are variable and depend primarily on the size of the protected data, the daily recovery point size, the expected volume data growth rate, and retention range objectives.

Daily recovery point size refers to the total size of changes that are made to protected data during a single day. This is roughly equivalent to the size of an incremental backup. Retention range refers to the number of days for which you want to store recovery points of protected data on disk. For files, DPM can store a maximum of 64 recovery points for each volume that’s included in a protection group, and it can create a maximum of 8 scheduled recovery points for each protection group each day.

Note

The limit of 64 recovery points for files is a result of the limitations of the Volume Shadow Copy Service (VSS), which is necessary for end-user recovery functionality of DPM. The recovery point limit does not apply to application data.

In general, we recommend making the storage pool three times the size of the protected data for protection of files. This recommendation is based on an assumed daily recovery point size of approximately 10 percent of the protected data size and a retention range of 10 days (two weeks, excluding weekends).

Note

The best way to calculate for specific workloads is to use the DPM storage calculator. Click Storage calculators for System Center Data Protection Manager 2010 to download the calculator.

If your daily recovery point size is larger or smaller than 10 percent of your protected data size, or if your retention range objectives are longer or shorter than 10 days, you can adjust the capacity requirements for your storage pool accordingly.

Regardless of how much capacity you decide to allow for the storage pool in your initial deployment, we recommend that you use extensible hardware so that you have the option of adding capacity if the need arises.

The following sections provide guidelines for determining your daily recovery point size and retention range objectives.

Estimating daily recovery point size

Our recommendation to make the storage pool three times the size of the protected data assumes a daily recovery point size of 10 percent of the protected data size. Daily recovery point size is related to data change rate and refers to the total size of all recovery points created during a single day. To get an estimate of the daily recovery point size for your protected data, you can review an incremental backup for a recent average day. The size of the incremental backup is usually indicative of the daily recovery point size. For example, if the incremental backup for 100 GB of data includes 10 GB of data, your daily recovery point size will probably be about 10 GB.

Determining retention range objectives

Our recommendation to make the storage pool three times the size of the protected data assumes a retention range objective of 10 days (two weeks, excluding weekends). A retention range of 10 days provides for recovery of data up to two weeks after a data loss event.

The longer your retention range objective, the fewer recovery points you can create each day. For example, if your retention range objective is 64 days, you can create just one recovery point each day. If your retention range objective is 8 days, you can create eight recovery points each day. With a retention range objective of 10 days, you can create approximately six recovery points each day.

Calculate DPM storage

One of the biggest challenges in DPM is calculating the size of the disks for storage pool that are used for protection of data sources. In general, we use the total size of the data that is to be protected (known as a data source) and multiply it by 3 times. It may look like the following:

DS x 3 =

There are more complex calculations for specific workloads that can be done based on data types. (i.e.: Hyper-V, SharePoint, SQL, Exchange and so on) For example, assume that the data source size is 4TB:

DS 4TB x 3 = 12TB

Note The best way to calculate for specific workloads is to use the DPM storage calculator. To download the DPM storage calculator, click: here.

The next step is to know your data's current growth rate. We will use a 7% growth rate as an example. If you take the current 4TB and added a 7% growth rate onto this, it would be .28 for a year. After one year, it would be 4.28TB.

4TB x (1 + 7%) = 4.28TB

After taking the growth rate into account you can see that the 12TB would cover the storage needs for a while at that growth rate. You also could calculate the storage including the growth rate.

Note This may change a bit depending on what you have for specific data types. Generally, we will do storage calculations like this to start with and grow it as needed. With DPM you can add more storage later on as needed.

For monthly and yearly storage those would be sent to tape and or Azure backup. DPM cannot do long term storage to disk.

Plan the disk configuration

If you are using direct-attached storage for the DPM storage pool, you can use any hardware-based configuration of redundant array of independent disks (RAID), or you can use a "just a bunch of disks" (JBOD) configuration.

Note Do not create a software-based RAID configuration on disks that you will add to the storage pool. Software-based RAID will make it very difficult to migrate to new disks as you cannot move software based raid volumes between disks. Also, DPM should be a dedicated single purpose system and should not be used to host non-DPM related data, so there should be no need to create software based RAID volumes on DPM managed disks.

To decide on the configuration for the disks, consider the relative importance of capacity, cost, reliability, and performance in your environment. For example, because JBOD does not consume disk space for storing parity data, a JBOD configuration makes maximum use of storage capacity. For the same reason, the reliability of JBOD configurations is poor; a single disk failure inevitably results in data loss.

For the typical DPM deployment, DPM recommends a RAID 5 configuration, which offers an effective compromise among capacity, cost, reliability, and performance.

To help you evaluate options for configuring the disks in your storage pool, the following table compares the trade-offs between JBOD and the various levels of RAID, on a scale from 4 (very good) to 1 (acceptable).

Disk Configuration Capacity Cost Reliability Performance and Scalability
JBOD 4 4 1 4
RAID 0 4 4 1 4
RAID 1 1 1 4 3
RAID 5 3 3 3 3
RAID 10 1 1 4 4

For more information about RAID, see Achieving fault tolerance by using RAID.

Define custom volumes

In System Center 2012 – Data Protection Manager (DPM), you can assign a custom volume to a protection group member, instead of leting DPM create them for you automatically in the storage pool. A custom volume is a dedicated volume created manually and formatted NTFS to be used as a replica or recovery point volume when adding a data source to protection. This custom volume can reside on any physical disk on the DPM server including disks used for DPM storage pool. To do this, follow these steps:

  1. Run the Create New Protection Group Wizard or the Modify Protection Group Wizard.

  2. In the Review Disk Allocation window, click Modify.

  3. In the Modify Disk Allocation window, click the drop-down menu under Storage Type.

  4. Select Custom Volume, and then finish the configurations that you want.

  5. Click OK, and finish the steps in the wizard.

Although the DPM-managed storage pool is sufficient for most business needs, you might want a greater degree of control over storage for specific data sources. For example, you may have critical data that you want to store by using a high-performance logical unit number (LUN) on a storage area network.

Any volume that's attached to the DPM server can be selected as a custom volume in the Create New Protection Group Wizard except for the volume that contains the system and program files. To use custom volumes for a protection group member, two custom volumes must be available: one to store the replica and one to store the recovery points.

DPM cannot manage the space in custom volumes. Custom volumes in DPM do not use the Autogrow feature. If DPM alerts you that a custom replica volume or recovery point volume is running out of space, you must manually change the size of the custom volume in Disk Management.

You cannot change the selection of using the DPM storage pool or custom volume for a protection group member after the group is created using the DPM GUI. If you must change the storage location for a data source's replica or recovery points, you can do so using the MigrateDatasourceFromDpm.ps1 powershell script.