Administrator's Guide


Client Operations Controlled by Policy

ADSM policies govern the following client operations, which are discussed in this section:

Backup and Restore

ADSM clients can back up and restore files and directories.

Backup

To guard against the loss of information, ADSM can copy files, subdirectories, and directories to media controlled by ADSM. Backups can be controlled by administrator-defined policies and schedules, or users can request backups of their own data. ADSM provides two types of backup:

Incremental backup
The backup of files according to policy defined in the backup copy group of the management class for the files. An incremental backup typically backs up all files that are new or that have changed since the last incremental backup.

Selective backup
Backs up only files that the user specifies. The files must also meet some of the policy requirements defined in the backup copy group.

Restore

When a user restores a backup version of a file, ADSM sends a copy of the file to the client node. The backup version remains in ADSM storage.

If more than one backup version exists, a user can restore the active backup version of the file or any inactive backup versions.

If policy is properly set up, a user can restore backed-up files to a specific time. See "Defining and Updating a Backup Copy Group" for details on the requirements.

Archive and Retrieve

To preserve files for later use or for records, a user can request ADSM to copy files, subdirectories, and directories for long-term storage on media controlled by ADSM. When users archive files, they can choose to have ADSM erase the original files from their workstation after the files are archived.

When a user retrieves a file, ADSM sends a copy of the file to the client node. The archived file remains in ADSM storage.

Client Migration and Recall

If the Hierarchical Storage Management (HSM) feature of ADSM is activated on a client node, users can migrate files from client node storage to server storage and recall files to the client node as needed. HSM frees space on client nodes for new data and makes more efficient use of your storage.

Files that are migrated and recalled with the HSM client are also called space-managed files.

For details about using HSM on clients, see ADSM Using the UNIX HSM Clients.

Migration

When a file is migrated to the server, it is replaced on the client node with a small stub file of the same name as the original file. The stub file contains data needed to locate the migrated file on server storage.

ADSM provides selective and automatic migration. Selective migration lets users migrate files by name. The two types of automatic migration are:

Threshold
If space usage exceeds a high threshold set at the client node, migration begins and continues until usage drops to the low threshold also set at the client node.

Demand
If an out-of-space condition occurs for a client node, migration begins and continues until usage drops to the low threshold.

To prepare for efficient automatic migration, ADSM copies a percentage of user files from the client node to the server. The premigration process occurs whenever ADSM completes an automatic migration. The next time free space is needed at the client node, the files that have been premigrated to the server can quickly be changed to stub files on the client. The default premigration percentage is the difference between the high and low thresholds.

Files are selected for automatic migration and premigration based on the number of days since the file was last accessed and also on other factors set at the client node.

Recall

ADSM provides selective and transparent recall. Selective recall lets users recall files by name. Transparent recall occurs automatically when a user accesses a migrated file.

Reconciliation

Migration and premigration can create inconsistencies between client node and server storage. For example, if a user deletes a migrated file from the client node, the copy remains at the server. At regular intervals set at the client node, ADSM compares client node and server storage and reconciles the two by deleting from the server any outdated files or files that do not exist at the client node.


[ Top of Page | Previous Page | Next Page | Table of Contents | Index ]