Physical Sequential Transparency › Implementation Considerations › Automated Cleanup of Uncataloged Data Sets

Automated Cleanup of Uncataloged Data Sets

Unlike VSAM data sets, Physical Sequential data sets are often defined by GDGs or similar schemes - instead of remaining constant, the data set name for a given application keeps changing as new generations are created and old ones fall away. This can easily lead to uncontrolled growth of the Control File, so the CFMU provides the ERASEUNCAT keyword to specify that a PS dsname entry should be automatically deleted from the Control File after it is uncataloged. To avoid performance problems and to prevent entries from being lost due to inadvertent deletion of the data set, the Control File is not purged until at least the day after the data set is uncataloged.

For your convenience, CA Compress tries to pick an intelligent default. If you define a PS pattern with GDG=ONLY, CA Compress realizes that only GDG data sets are permitted and selects ERASEUNCAT=YES. For GDG=NO or GDG=YES, which permit non-GDG data sets, CA Compress defaults to ERASEUNCAT=NO for safety. In many cases, however, such as IMS log data sets, you probably want to override this choice with ERASEUNCAT=YES.