Your Browser is not longer supported

Please use Google Chrome, Mozilla Firefox or Microsoft Edge to view the page correctly
Loading...

{{viewport.spaceProperty.prod}}

Pubset backup with Snapsets

A Snapset is the backup copy of an SF or SM pubsets. A Snapset serves as the pubset backup for restoring lost data (e.g. data deleted inadvertently).

Snapsets are created by systems support or the HSMS administrator during ongoing pubset operation and are deleted again later. The oldest Snapset can be implicitly deleted during the generation process.
The Snapsets can be used as a logical backup of all files and job variables of a pubset. The complete pubset can also be reset to the status at the time the Snapset was created.

Snapsets are placed in service when they are generated or when the pubset is imported and can then be accessed in read mode. They permit individual files and job variables to be restored to the status of the Snapset involved.

Snapsets can also be used in shared pubset mode.

Snapsets of a Pubset are not availiable temporary, if the subsystem SHC-OSD was not active at the moment of the pubset imports (particularly the snapsets of the home pubsets are not put automatically into operation when starting the system). As soon as SHC-OSD is active and the command SHOW-SNAPSET-CONFIGURATION is called, these snapsets are supplementaryput into operation .

Files on the Snapset

The CHECK-SNAPSET-CONFIGURATION command is used to check or activate the Snapset configuration of an imported pubset.

The LIST-FILE-FROM-SNAPSET command or LFFSNAP macro enables users to obtain information about files which have been backed up on a Snapset. They can obtain information about all files which they can access (as with SHOW-FILE-ATTRIBUTES, which provides information from the current file catalog).

Files backed up in this way can be restored from the Snapset using the RESTORE-FILE-FROM-SNAPSET command or RFFSNAP macro. During restoration, individual files are copied from the Snapset to the active pubset. This operation is comparable to an HSMS restore from a backup archive.

A particular backup version (the latest Snapset backup is preset) can be specified. Alternately, the latest version of the files can be restored on the basis of all existing Snapsets.

All file attributes of a restored file are taken over unchanged from the original file (including the creation and change dates and the protection attributes). Only the allocation can differ from the original file, also in the case of files with physical allocation. Files on SM pubsets are restored on the “most suitable” volume set. This need not be the same as the original volume set.

Individual file generations can only be restored with the entire file generation group. Files on private disk are not taken into account. In the case of migrated files and tape files, only the catalog entries are restored (without checking the availability of the associated tapes). Migrated files and tape files cannot be renamed when they are restored.

A nonprivileged user can restore a file from a foreign user ID only if they have read permission for the original file and are the owner of the target file.

Overwriting through restoration must be explicitly permitted for existing files (REPLACE operand). For files which are protected against unauthorized overwriting by means of a password, the required password must be entered in the caller’s password table (see the ADD-PASSWORD).

Files can also be restored under a new name (NEW-FILENAME operand). They are renamed by specifying either a different user ID or a file name prefix.

Files which were opened for writing when the Snapset was created can optionally be restored (RESTORE-OPEN-FILES operand). A file restored in this way has the same status as when the system crashed. Verification may be required for an ISAM file. When a Snapshot is created, the system files of SRPM, Guards and GCF, which are permanently open, are placed in a consistent status which permits them to be restored later.

If required, the caller can have a log of restore processing output to SYSOUT or SYSLST (OUTPUT operand). The log can cover either all files or only those files which, for particular reasons, could not be restored (REPORTING operand).

Job variables on the Snapset

The same mechanism is used for job variables as for files (see the section above).

The only differences are in the asasociated commands and/or operands:

  • The LIST-JV-FROM-SNAPSET operand or LJFSNAP macro provides information about job variables which have been backed up on a Snapset (as with SHOW-JV-ATTRIBUTES, which provides information about job variables from the current file catalog).

  • The RESTORE-JV-FROM-SNAPSET command or RJFSNAP macro restores job variables to the status of a particular Snapset or to the latest status on the basis of all existing Snapsets. Renaming takes place here using the NEW-JV-NAME operand. The other options for overwriting (REPLACE operand) and controlling log output (OUTPUT and REPORTING operands) are the same as for restoring files.

Displaying Snapsets

Users can obtain information on existing Snapsets (= existing backup versions of the pubset) using the SHOW-SNAPSET-CONFIGURATION command.

Output takes place to SYSOUT, or optionally also to SYSLST.
The creation date, the Snapset ID, the ID of the CCOPY session via which the Snapset can be accessed, the name of the assigned snap pool and the ID of the SRDF target disk storage system (among other things) can be output for each Snapset if snap copies were also created for the target units (see also the “SHC-OSD” manual [26 (Related publications)]).

Restoring library members

While a PLAM library is restored as a complete file using the RESTORE-FILE-FROM-SNAPSET command or the RFFSNAP macro, the PLAM library in LMS can be opened directly on the Snapset. This permits read access to the individual library members and consequently restoration on a member-by-member basis.

Incorporating files and job variables from Snapsets in the backup archive

The HSMS statement BACKUP-FILES enables files and job variables which have been backed up on a Snapset to be incorporated in a backup archive. As a result, files and job variables can practically be transferred “offline” to active pubset operation for long-term backup. For details, please see the “HSMS” manual [10 (Related publications)].