Create the ABR Catalog

New installations

If you have ABR, you must create the ABR catalog.

Existing installations

You should use the existing ABR catalog.

The ABR catalog

ABR needs to save pointers to many of the backups it creates. Rather than creating and managing an external database for this information, ABR stores this information in a standard ICF user catalog. In addition, ABR needs to save backup information for data sets that have been scratched from their DASD volumes.

Each type of entry must be in a single ICF catalog with a one-level alias; ABR does not support Multi-Level Alias (MLA) for the ABR catalog. The two types of ABR catalog entries can be in two separate catalogs, but most installations put them in a common ABR catalog. You can even put other types of catalog entries in the same catalog or assign the ABR aliases to an existing catalog, but these are strongly discouraged.

If the ABR catalog must ever be expanded or moved, member MOVECAT in the FDR FDRSAMP installation library contains instructions and guidance.

Backup section

The backup section of the ABR catalog contains an entry for every full-volume, incremental, and selective backup file created by ABR (files created by running FDRABR with DUMP TYPE=FDR, AUTO, ABR, or DSF - Working-with-FDRABR-Volume-Backups). These entries normally begin with “FDRABR.”. ABR automatically maintains this section, deleting old generations as new generations are created, retaining the number of generations specified for each volume (see Set-ABR-Disk-Volume-Processing-Options) so backup entries are usually retained only for a month or so.

Scratch section

The scratch section of the ABR catalog contains an entry for data sets that were deleted from DASD after having been backed up by ABR (full-volume or incremental backup). These entries are only created if the ABR DADSM Preprocessing exit is installed (DADSM Pre-Processing Exit in Dynamic-Installation-of-the-FDR-Exits). The exit prefixes a high-level index of “#.” to the name of the data set being scratched, and catalogs the resulting name in the ABR catalog. These entries identify the backup file if a user requests ABR to restore a scratched data set. The PURGE SCRATCH command of program FDRABRCM (Catalog-Maintenance-Utility-FDRABRCM) deletes these entries when the backups with which they are associated are no longer in the catalog; this must be run periodically to keep the catalog from filling up.

Archive section

The archive section of the ABR catalog contains entries for certain ARCHIVE backups (DUMP TYPE=ARC – see Working-with-FDRABR-Archiving-and-Superscratch) and sometimes application backups (DUMP TYPE=APPL – see Working-with-FDRAPPL). These entries normally begin with “FDRABR.”. ARCHIVE normally catalogs only the first backup created in each step and any backup that starts a new reel of tape; application backups normally do not catalog any backups. However, if requested (ARCCAT=ALL or EXPDT=99000 in the JCL) ABR catalogs all those backups. Backups on DASD are always cataloged. These backups are also recorded in an Archive Control File (ACF) but the catalog entries may also be used for restores. The REORG command of program FDRARCH (FDRARCH-REORG-Statement) removes these entries when the reorganization process finds that the archive tape has expired.

Catalog aliases

The ABR catalog must be assigned catalog aliases (in the z/OS master catalog) of:

FDRABR

for the backup and archive entries

#

for the scratch entries

If you have changed the high-level indexes used for these entries in the FDR Global Options (see ABR-Options - panel A.I.4.5), use the values you have assigned instead.

Panel A.I.7 – Create the ABR Catalog Panel

Select option 7 (ABRCAT) on the FDR Installation Options Menu that will bring you to this panel, which creates IDCAMS batch JCL for the allocation of the catalog.

Create the ABR Catalog – Panel A.I.7

ABR Catalog Name

Specify the name you want to assign to the ABR catalog. The BMC recommendation “CATALOG.ABRBASE” is shown as a default but you may change it to anything that meets your installation’s requirements.

The master catalog name can be omitted unless the ABR catalog must be defined in the master catalog of a system that is not the system the job will execute on. The master catalog named must be defined as a user catalog on the system where the job will run.

Catalog specifications

This section describes the characteristics of the catalog to be allocated.

If the ABR catalog is shared by multiple z/OS systems that are not using a shared master catalog, it must be defined in the master catalog of each of those systems, so specify “YES” for SHARED CATALOG. This takes you to another panel where you provide the master catalog name for each of those systems (each of them must be defined as a user catalog on the system where the job will run). If the master catalog is shared, specify “NO”.

The remainder of the parameters specify where to allocate the catalog and how big to make it. According to estimates in the IBM manual Managing Catalogs, about 200 bytes are required for a non-VSAM entry, which translates to about 3500 entries in a 3390 cylinder. From the descriptions earlier in this section, you can estimate how many entries will exist in the catalog, but we recommend that you generously over allocate the catalog to allow for expansion and free space. An allocation of 25 to 50 cylinders is adequate in most installations.

ABR indexes

These fields are filled in with the values for the ABRINDEX and SCRINDEX in the FDR Global Options; there should be no reason to override them.

Submit the job

Update the job information at the bottom of the screen and then enter, on the COMMAND line at the top, either “SUBMIT” to immediately submit the job or “EDIT” to review, edit, and submit it manually (or save it).