Plan for protection groups

 

Updated: May 13, 2016

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

A DPM protection group is a collection of data sources, such as volumes, shares, or application workloads which have common backup and restore settings. The protection group settings specify:

  • Data sources—The servers, computers, and workloads you want to protect

  • Back-up storage—How the protected data should be backed up in the short-term and long-term.

  • Recovery points—The recovery points from which replicated data can be recovered.

  • Allocated disk space—The disk space that will be allocated to data in the protection group from the storage pool.

  • Initial replication—How the initial replication of data should be handled, using either over the network or manually offline.

  • Consistency checks—How replicated data should be checked for consistency

  • Which data sources will belong to the protection group?

  • Which protection method (disk-based, tape-based, or both) will you use for the protection group?

  • What are your recovery goals for the members of the protection group?

  • How much storage space will be needed to protect the selected data?

  • Which tape and library should be used?

  • What method will you use to create the replica for the members of the protection group?

The topics in this section provide guidelines for making the decisions involved in creating a protection group.

Planning protection groups

You can define your protection groups in a number of ways:

  • By computer—So that all data sources for a computer belonging to the same protection group. This provides a single point of adjustment for the computer’s performance loads. However, all data sources will then have the same backup and recovery settings.

  • By workload—So that you separate files and each application data type into different protection groups. This allows you to manage workloads as a group. However recovering a multi-application server might require multiple tapes from different protection groups.

  • By RPO/RTO—Gather data sources with similar Recovery Point Objectives (RPOs) and Recovery Time Objectives (RTOs). You control the RPO by setting the synchronization frequency for the protection group which determines the amount of potential data loss (in time) in the case of unexpected outages. The RTO measures the acceptable amount of time that data is unavailable and is affected by the storage methods your select for the protection group.

  • By data characteristics—For example how often data changes, how rapidly it grows, or its storage requirements.

Note the following:

  • If you have only a single stand-alone tape, use a single protection group to minimize the effort to change tapes. Multiple protection groups require a separate tape for each protection group.

  • Data sources on a computer must be protected by the same DPM server. In DPM, a data source is a volume, share, database, or storage group that is a member of a protection group.

  • You can include data sources from more than one computer in a protection group.

  • Protection group members cannot be moved between protection groups. If you decide later that a protection group member needs to be in a different protection group, you must remove the member from its protection group and then add it to a different protection group.

  • If you determine that the members of a protection group no longer require protection, you can stop protection of the protection group. When you stop protection, your options are to retain protected data or to delete protected data.

    • Retain protected data option: Retains the replica on disk with associated recovery points and tapes for the specified retention range.

    • Delete protected data option: Deletes the replica on disk and expires data on the tapes.

  • When you select a parent folder or share, its subfolders are automatically selected. You can designate subfolders for exclusion and also exclude file types by extension.

  • Verify that you do not have more than a 100 protectable data sources on a single volume. If you do, distribute your data sources across more volumes if possible.

  • When you select a data source that contains a reparse point (mount points and junction points are data sources that contain reparse points), DPM prompts you to specify whether you want to include the target of the reparse point in the protection group. The reparse point itself is not replicated; you must manually re-create the reparse point when you recover the data.

Set performance options

Throttling— Network bandwidth usage throttling is configured at the protected-computer level. In addition, you can specify different network bandwidth usage throttling rates for work hours, non-work hours, and weekends, and you define the times for each of those categories.

Compression—On-the-wire compression is configured at the protection-group level for backup to tape. Compressing data reduces the space needed on the tape and increases the number of backup jobs that can be stored on the same tape. Compression doesn’t significantly increase the tme required to complete the backup job. Encryption increases data security, and also doesn’t significantly increase the time required for the backup job. Encryption requires a valid certificate on the DPM server.

Reduced scheduling—When protecting application data such as Exchange storage groups or SQL Server databases over a WAN, consider reducing the schedule for express full backups.

Allocate storage space

When you create a protection group and select disk-based protection, you must allocate space on the storage pool for the replicas and recovery points for each data source that you have selected for membership in the group, and you must allocate space on protected file servers or workstations for the change journal.

DPM provides default space allocations for the members of the protection group. The following table shows how DPM calculates the default allocations.

How DPM Calculates Default Space Allocations

Component Default Allocation Location
Replica volume For files:

- (Data source size x 3) / 2

For Exchange data:

- Data source size x (1 + log change) / (alert threshold - .05)

For SQL Server data:

- Data source size x (1 + log change) / (alert threshold - .05)

For Windows SharePoint Services data:

- Total size of all databases/ (alert threshold - .05)

For Virtual Server data:

- Data source size x 1.5

For system state:

- (Data source size x 3) / 2

For Hyper-V

- Data source size x 1.5
DPM storage pool or custom volume
Recovery point volume For files:

- (Data source size x retention range in days x 2) / 100 + 1600 MB

For Exchange data:

- 4.0 x retention range in days x log change x data source size + 1600 MB

For SQL Server data:

- 2.5 x retention range in days x log change x data source size + 1600 MB

For Windows SharePoint Services data:

- 1.5 x retention range in days x log change x total size of all databases + 1600 MB

For Virtual Server data:

- (Data source size x retention range in days x 0.02) + 1600 MB

For system state:

- (Data source size x retention range in days x 2) / 100 + 1600 MB

For Hyper-V

- (Data source size * recovery range in days * 0.1) + 1600 MB
DPM storage pool or custom volume
Change journal (for file protection only) 300 MB Protected volume on the file server or workstation

To help you figure out storage capacity use HYPERLINK "http://go.microsoft.com/fwlink/?LinkId=275371" Storage Calculators for DPM

The values used in the preceding table are defined as follows:

  • **Alert%—**Threshold for the alert associated with replica growth; typically 90%.

  • **Log change—**This is the change rate on the database or storage group in question. This varies widely, but for the purposes of the default recommendation in DPM, it is assumed to be 6% for Exchange and SQL Server data and 10% for Windows SharePoint Services data.

  • **Retention range (RR)—**This is the number of recovery points stored; it is assumed to be 5 for purposes of the DPM default recommendation.

  • **System state data source size—**The data source size is assumed to be 1 GB.

When you create a protection group, in the Modify Disk Allocation dialog box, the Data Size column for each data source displays a Calculate link. For the initial disk allocation, DPM applies the default formulas to the size of the volume on which the data source is located. To apply the formula to the actual size of the selected data source, click the Calculate link. DPM will determine the size of the data source and recalculate the disk allocation for the recovery point and replica volumes for that data source. This operation can take several minutes to perform.

We recommend that you accept the default space allocations unless you are certain that they do not meet your needs. Overriding the default allocations can result in allocation of too little or too much space.

Allocation of too little space for the recovery points can prevent DPM from storing enough recovery points to meet your retention range objectives. Allocation of too much space wastes disk capacity.

If, after you have created a protection group, you discover that you have allocated too little space for a data source in the protection group, you can increase the allocations for the replica and recovery point volumes for each data source.

If you discover that you have allocated too much space for the protection group, the only way to decrease allocations for a data source is to remove the data source from the protection group, delete the replica, and then add the data source back to the protection group with smaller allocations.

To help you estimate your storage space needs, download the DPM storage calculator.

Initial replication

When you create a protection group, you must choose a method for creating the initial replica, which copies all the data selected for protection to the DPM server and then runs synchronization with consistency check for each of the replicas.

Initial replication over the network

DPM can create the replicas automatically over the network, or you can create the replicas manually by restoring the data from removable media such as tape. Automatic replica creation is easier, but, depending on the size of the protected data and the speed of the network, manual replica creation can be faster.

To help you choose a replica creation method, the followingtable provides estimates for how long DPM takes to create a replica automatically over the network given different protected data sizes and network speeds. The estimates assume that the network is running at full speed and that other workloads are not competing for bandwidth. Times are shown in hours.

Hours to Complete Automatic Replica Creation at Different Network Speeds

Size of Protected Data 512 Kbps 2 Mbps 8 Mbps 32 Mbps 100 Mbps
1 GB 6 1.5 < 1 < 1 < 1
50 GB 284 71 18 5 1.5
200 GB 1137 284 71 18 6
500 GB 2844 711 178 45 15

Initial replication manually

If you are deploying DPM to protect data over a WAN and your protection group includes more than 5 GB of data, we recommend that you choose the manual method for creating the replicas.

If you choose manual replica creation, DPM specifies the precise locations on the DPM server where you must create the replicas. Typically, you create the replicas by restoring your most recent backup of the data source from removable media such as tape. After you restore the data, you complete the process by running synchronization with consistency check for each of the replicas.

It is crucial that when you restore the data to the DPM server to create the replica, you retain the original directory structure and properties of the data source, such as time stamps and security permissions. The more discrepancies that exist between the replicas and the protected data source, the longer the consistency checking part of the process takes. If you do not preserve the original directory structure and properties, manual replica creation can take as long as automatic replica creation.