From FILEPOOL BACKUP

Follow these steps to restore a user storage group that was backed up by the FILEPOOL BACKUP command:

Inform file pool users of the impending restore. Users need to be aware of restore because the storage group data will regress to some earlier state. Furthermore, the storage group will not be available for read or write access during the restore. You should advise file pool users against attempting to reference the storage group until the restore completes. (If they ignore your advice, no harm will be done to the file pool. Their applications or commands, however, may get damaged return codes.)
Users enrolled in the storage group being restored will not be able to connect to the file pool until the restore completes successfully.
If you intend to stop the server before restoring user storage groups, consider increasing the CATBUFFERS startup parameter value to 5000 in the DMSPARMS file. Increasing CATBUFFERS prior to restoring a user storage group should reduce the amount of time it takes to do the restore.
Ensure the file pool is available in multiple user mode.
Log on a virtual machine that has file pool administration authority.
The administrator that enters the FILEPOOL BACKUP command must be on the same processor as the file pool. The FILEPOOL BACKUP command links storage group minidisks to the administration machine, and the CP LINK command cannot link minidisks from a remote processor.

For the VMSYS, VMPSFS, and VMSYSU file pools, the MAINT and MAINTvrm virtual machines have administration authority.
Enter a FILEDEF command (and, optionally, a LABELDEF command) for ddname RESTORE to identify the backup file.
Suppose, for example, you want to replace the contents of storage group 3. The backup file is the second file on tape volume SYSTST, which uses IBM standard labels. When making the backup, you used a LABELDEF command that specified STORAGE3 for the FID operand and 2 for the FSEQ operand. Assuming the tape is mounted at virtual device number 181, you would enter:
```
filedef restore tap1 sl 2 volid systst
labeldef restore fid storage3 fseq 2
```
When you enter the FILEPOOL RESTORE command in a later step, it will use the LABELDEF information to verify the tape label.

For more on the FILEDEF and LABELDEF commands, see z/VM: CMS Commands and Utilities Reference. For more about using tapes with CMS, see z/VM: CMS Application Development Guide for Assembler.
If the backup file is a CMS file named SG3 DATA that resides on a minidisk or SFS directory accessed as file mode B, you would enter:
```
filedef restore disk sg3 data b
```
The file you identify should be the most current backup. If, for some reason, you are not using the most current backup, you should read the usage note concerning Obsolete Backup Files listed under the description of FILEPOOL RESTORE.
Optionally enter a SET FILEWAIT ON command:
```
set filewait on
```
If you do not enter the SET FILEWAIT command, the FILEPOOL RESTORE command will not work if anyone is reading from or writing to the storage group. If you enter the command, the FILEPOOL RESTORE command will wait instead.

For more information on the SET FILEWAIT command, see z/VM: CMS Commands and Utilities Reference.
Ensure file mode A is accessed as read/write and that it is not in the storage group being restored. This is needed because the FILEPOOL RESTORE processing uses the CMS GLOBALV facility to keep track of certain information needed for recovery from catastrophic error. Catastrophic error is any error that causes the usual stopping logic to be avoided.
Enter the FILEPOOL RESTORE command. An abbreviated format of the command is:
To restore storage group 3 of the VMSYSU file pool, for example, you would enter:
```
filepool restore 3 vmsysu
```
During the restore, the storage group will not be available to other users. After the restore completes, users can use the storage group immediately. You should send the storage group users a message indicating the restore is complete. If you would prefer to inhibit multiple user mode access until you have had a chance to verify all went well, see Restricting User Access in Multiple User Mode.
At the end of its processing, the FILEPOOL RESTORE command displays the following message:
```
DMS3500I Restore of storage group nn
in file pool filepoolid successfully completed
```
If the FILEPOOL RESTORE command is unsuccessful before the stopping message is displayed, you should immediately rerun it. Otherwise, certain resources may be left in an unusable state. The storage group itself is disabled early in FILEPOOL RESTORE processing. (That is, FILEPOOL RESTORE processing gets an exclusive disable lock on the storage group.) You should never, in this case, manually enable the storage group. The catalog data for the storage group may be corrupt. Re-enabling the storage group could lead to unpredictable results. Instead, run the FILEPOOL RESTORE command again.

If the rerun of the FILEPOOL RESTORE command does not succeed, you have to enter the FILEPOOL CLEANUP command to make available as many resources as possible.

For more information, see FILEPOOL RESTORE and FILEPOOL CLEANUP.

Note: If you are restoring more than one storage group in the same file pool, do not try to restore them at the same time. Because the FILEPOOL RESTORE command updates the catalog data, multiple FILEPOOL RESTORE commands could interfere with each other, causing them to not complete. Always restore storage groups one at a time.
Enter a SET FILEWAIT OFF command.
```
set filewait off
```
You need to enter this command only if you entered SET FILEWAIT ON in a preceding step.
If you increased the CATBUFFERS parameter in the DMSPARMS file, you should reset the CATBUFFERS value at the earliest opportunity.
Determine whether aliases are intact. See Considerations for SFS Aliases After FILEPOOL FILELOAD for details.