Policy concepts

The policy for a specific client is determined by the settings in the policy domain to which a client is added.

During the client registration process, you assign a client to a policy domain. The policy for each client is determined by the rules in the policy domain to which the client is assigned. In the policy domain, the rules that are in effect are in the active policy set.

When a client backs up or archives a file, the file is bound to a management class in the active policy set of the policy domain. A management class is the key set of rules for managing client data. The backup and archive operations on the client use the settings in the default management class of the policy domain unless you customize policy.

A policy can be customized by defining more management classes in the policy set, activating the policy set, and assigning the use of the new management classes through client options.

Client options can be specified in a local, editable file on the client system and in a client option set on the server. The options in the client option set on the server can override or add to the options in the local client option file.

The server uses the policy in management classes to manage files based on whether file versions are active or inactive. The most recent backup or archived copy of a file is the active version. Active versions are never deleted from server storage.

Backup versions other than the most recent version are called inactive versions. An active version of a file becomes inactive when one of the following events occur:

The file is backed up again, creating a more recent version of the file in server storage.
The file is deleted from storage on the client node and then an incremental backup operation runs. An incremental backup, the typical backup operation for a client, backs up only those files that changed since the last backup.

The settings in the management class that is bound to a file determine how long and how many inactive versions of the file are retained.

Expiration processing uses policies to determine when inactive versions are no longer needed, that is, when the versions are expired. The process of expiration on the server enforces policies that you define for data retention, and you must ensure that you schedule expiration to run regularly. For example, if you have a policy that requires only four versions of a file be kept, the fifth and oldest version is expired. During expiration processing, the server removes entries for expired versions from the database, which in effect deletes the versions from server storage.