Understanding Tape Retention › Retention Master File (RMF)

Retention Master File (RMF)

The Retention Master File (RMF) maintains the retention instructions for your tape data sets. These instructions are supplied by you in the form of control statements which create records in the RMF.

• Retention control statements allow you to define different retention periods for data sets that require unique handling. You can code the retention statements to include individual data sets or groups of data sets by using data set name qualifiers.
• A default retention statement allows you to define your data center standard retention period to those data sets not covered by specific retention requirements.
• Location control statements permit you to define movement of data sets from one storage location to another. These statements describe the location, the kind of storage that will be used there (cabinet/slot, box or undefined), and the maximum number of tapes that can be stored. When you define retention for your critical data sets, you can specify that they be moved to an off-site storage location at the end of their data center retention period. When the retention period ends at the off-site location, you can specify that they then be moved to other locations. When no other location is specified, the tapes will be returned to the data center for reuse as scratch tapes.

After you have defined this movement, the Tape Retention System (TRS) automates its implementation, deciding when and where to move the tapes, recording their location so they can be found when needed, telling you when to move the tapes to another location and when to return them to the data center for reuse.

You may also use the TRS to control movement of External Data Manager (EDM) volumes. EDM tapes may be processed separately or in a group with other tapes. Since CA TLMS will not scratch an EDM owned tape, each EDM is responsible for scratching tapes it no longer needs.

You can define different retention instructions for the data center and each off-site storage location. Each set of retention instructions may cover a specific data set, or a group of data sets with common data set name qualifiers.

TRS generates pseudo-data set names to allow flexible retention for tapes created with the Tape Stacking feature. Since the names of the data sets stacked on a tape may not be known, TRS generates a pseudo-DSN for each tape which has the stacking flag set. The DSNs will have the form:

CATAPE.STACKING.xxxx.poolname

where xxxx is PART or FULL, indicating whether the tape is partially or completely full. The last node is the eight character stacking scratch pool for the tape. The pseudo-DSN may be used like a normal DSN when coding retention rules. However, CATLG and CYCLE retentions should not be used.

TRS has no special processing for WORM (Write Once Read Many) tapes but CA makes some "best practices" recommendations. WORM tapes cannot be rewritten so the tapes cannot be physically reused as scratch tapes. CA TLMS does not prevent WORM tapes from being scratched but will automatically set the tape to "out-of-service" with an OUTCODE of "WORM." This will prevent CA TLMS from doing any further processing.

As a best practice, CA recommends that WORM tapes have a TRS schedule which does not scratch the tape. That is, the last location will have manual control. CA recommends a "Destroy WORM" location be added to the RMF and this location have a type 7 manual control. This should be the last location after the other rules where the tape is retained for as long as it is needed. When the other retentions are exhausted, the tape will be moved to the "destroy WORM" location. It is not intended that the WORM tape be retained at this final location. It is a signal that the tape can be destroyed without being scratched. After the tape is destroyed the VMF record can be cleared and a new WORM tape initialized with the same VOLSER.