General Information

This facility is provided not only to recover the dead space within a PDS, but also to enlarge or shrink the size of both the data set and its directory, based upon their content. The compression is accomplished by copying each data set to a work file, calculating the reallocation requirements, then copying it back. A message is issued each time the copy-back phase of a compress is begun.

Since a single work file is used for compressing all of the datasets, the size of the work file must be large enough to accommodate the compressed form of the largest data set to be encountered.

The work file can be on any DASD device type (it can span multiple volumes) and given any block size, since CA Disk restructures the data to satisfy the target work file.

Note: If you direct this work file to a specific volume serial (by overriding the COMPWRK DD statement with a VOL=SER parameter), COMPRESS will fail with a U200 abend. To avoid this abend, you must also include a DISP= parameter on your override statement.

In the event that a PDS that was copied to the work file cannot be reloaded due to an unsuccessful allocation or some other reason, message 3179, ACTIVE WORK FILE, CONTAINS dsname, will be issued and the job will terminate with a USER 200 abend. The work file will be kept to allow the data set to be recovered via the RELOAD command. The PDS can be reloaded to the original volume or to a new volume.

PDS compression can be accomplished implicitly by the designation of volumes to be scanned for eligible datasets, or explicitly by the specification of a data set name.

The report produced by PDS Compress contains two lines for each data set processed. The first line presents the status of the data set prior to compression. The second line shows its status after compression.