Organizing Protection Groups

  • Article

Published : April 8, 2005 | Updated : August 17, 2005

The principal factor to consider in deciding how to organize data into protection groups is tolerance for data loss; that is, how important it is to be able to quickly recover a recent copy of the data. Typically, you will want to gather data for which you have similar loss tolerance into the same protection group, and specify a protection schedule for the group that reflects that level of tolerance.

Other factors you may want to consider as you structure your protection groups include the business requirements of your organization, the departmental structure of your organization, the sensitivity of the data, the capacity of your network, geographic location of your file servers, and characteristics of the data, such as data type, rate of data change, and rate of data growth. Often you will want to gather data with similar characteristics into the same protection group.

 Considerations for protecting data over a WAN

Network bandwidth usage throttling and on-the-wire compression are performance optimization features that are particularly important for deployments in which a DPM server protects data over a WAN or other slow network. These features are enabled on protection groups. If you are deploying DPM to protect some data over a WAN or other slow network, and other data over a fast network, you may want to organize your protection groups around link speed. This allows you to adjust throttling and compression settings for the speed of each of your links.

As you design the structure of your protection groups, keep the following guidelines and restrictions in mind:

  • All protected data sources on the same volume must be protected in the same protection group.

    For example, if Volume A contains Folders 1 and 2, and you add Folder 1 to Protection Group X, you cannot add Folder 2 to any protection group other than Protection Group X.

  • A single protection group can include data sources stored on multiple file servers.

  • A single protection group can include multiple volumes, or data sources stored on multiple volumes.

  • Although you can include system volumes and program folders in protection groups, we recommend against including these data sources. Protecting a system volume does not enable you to restore the operating system or state of a file server. For instructions for protecting the system state of a file server, see “Protecting System State” earlier in this chapter.

    If a system volume contains user data that you want to protect, we recommend that you protect the relevant folders or shares individually rather than protecting the whole system volume.