Overview

Merge processing has been designed to provide maximum flexibility in managing the archive volumes on disk, tape and other storage media, and at the same time make the best use of the available resources.

When data sets are archived, each is assigned a retention period (expiration date). By definition, then, 100 percent of the data sets are unexpired on the day they are first archived. To make the most efficient use of both disk and tape, CA Disk (by default) packs all of these archived data sets into a single file (if archiving to tape), or a single data set allocated on disk.

As each day passes, however, some of these data sets can expire, creating dead space in the archives. Given enough time, all of the data sets will expire and the tape can be recycled or the archive data set on disk can be deleted to reclaim the space. (The IXMAINT function accomplishes this.)

In summary, each archive volume (tape or disk data set) starts out as packed full of 100 percent unexpired data sets. As the days pass, this will eventually drop to zero percent; that is, all data sets expired. The important questions are, (1) how fast does this take place and (2) can you afford to wait until it drops all the way to zero percent? In many cases if you do; tapes and disk space will be wasted. The archives need to be consolidated to reclaim the space.

This is the function that merge performs. Each merge run consolidates the unexpired data sets from one or more tape volumes (or archive data sets on disk) onto new archive volumes, thus freeing up tape volumes and/or disk space. This reduces the space required to store the CA Disk archives, as well as eliminating disk archive data sets with a large percentage of expired data sets. After the unexpired data sets have been successfully merged forward, the input data sets on disk are deleted, and input tapes are returned to scratch status — disk space is reclaimed and tapes are made available for reuse. All expired CA Disk index entries are dropped from the archives and unexpired entries are updated to reflect their locations on the new tape or disk archives.

Keeping some archives on disk can be desirable, because it maintains easy access to the data through auto-restore processing, without operator intervention. Disk‑to-disk merging provides an efficient means to manage these archives, also without operator intervention.