Saving Space on DFSMShsm-Owned Volumes

DFSMShsm can save space on migration volumes with its space saving functions. DFSMShsm generally moves and copies only used, not overallocated, space to its migration volumes. These DFSMShsm functions are available in system-managed and non-system-managed environments.

The following list represents the space saving functions:

Data compaction

DFSMShsm can save space on DFSMShsm-owned volumes by compacting each data set as it is processed. The data compaction functions are available with or without using the DFSMSdss component of DFSMS as the data set mover. DFSMShsm can compact data during data set backup and migration, or by using DFSMSdss compression during physical volume dumps and aggregate backup.

The IBM® RAMAC Virtual Array Subsystem also provides its own data compaction, making compaction by either DFSMShsm or DFSMSdss unnecessary.

Small data set packing

Small data set packing (SDSP) saves space on ML1 volumes by writing small data sets (fewer than 800 KB) as records in a VSAM data set called an SDSP data set. The space saving occurs because the DASD data set allocation requires allocation in full tracks. SDSP lets you give those data sets allocation in 2093-byte units.

Storage administrators can specify the size of data sets eligible to migrate to SDSP data sets in kilobytes or tracks. The storage administrator must specify which ML1 volumes have SDSP data sets.

Partitioned data set compression

Partitioned data set compression reduces the allocated space for a data set as members are deleted or replaced. DFSMShsm provides an automatic method of compressing partitioned data sets during space management by doing migration and recall when the partitioned data set exceeds a certain number of extents.

DFSMShsm maintains user information that is in partitioned data set directories when it performs partitioned data set compression.

User data set reblocking

During recall and recovery, the process of reblocking user data sets changes the number of records in a physical block and thereby uses the space on the DASD user volume more efficiently. Data movement using DFSMShsm reblocks physical sequential data sets. Data movement using DFSMSdss reblocks physical sequential and partitioned data sets.

Optimum DASD blocking

Optimum blocking of data is an option that saves space on DFSMShsm-owned DASD volumes by using an optimum block size for storing the maximum data on each track. The optimum block size is determined, based on the device type of the DASD device.

Tape recycle

Data on tapes is invalidated over time by the expiration or recall of migrated data sets or the generation of more recent backup data sets. DFSMShsm provides the capability of recycling backup or migration tapes when they contain less than an installation defined percentage of valid data. Recycling transfers the valid data sets from these tapes and consolidates the data onto fewer tapes, thus leaving the recycled tapes available for reuse. Tape recycle can also be used to move data from one tape technology type to another. Options are available to select input tapes by volume serial ranges, which can be used to identify tapes of a given technology.

Tapes eligible for recycle processing are logically sorted, with those containing the least amount of valid data (per cartridge type, standard or enhanced capacity) processing first.

To facilitate rapid recycle processing, up to 15 tape processing tasks can run simultaneously on a processor, with multiple tape buffers being used for improved throughput. The number of recycle tape processing tasks can be changed dynamically, even during recycle processing, to any number from one to 15.

Dump tapes and ABARS tapes can be reused automatically without recycle; dump tapes when they reach the end of their retention period, ABARS tapes when they reach an expiration date, or when aggregate roll off occurs.

If data in single-file format on a backup or ML2 tape is overwritten or becomes unreadable, DFSMShsm can retrieve all the undamaged data from that tape.