In an SMS environment, duplicate data set names are not allowed. Therefore, the following description applies only to a non-SMS environment.

Because duplicate data set names can cause confusion and possibly cause the wrong data set to be accessed, each data set, cataloged or uncataloged, should have a unique name. If you recover a cataloged data set but inadvertently include the RECOVER FROMVOLUME command and DFSMShsm backed up an uncataloged data set with the same name, DFSMShsm recovers the uncataloged version.

Another problem can occur with duplicate data set names when you access an uncataloged data set on multiple volumes and a migrated, cataloged, data set exists with the same name. In the JCL, you list the volume serial numbers of the volumes where that data set resides. If, however, the data set does not reside on one of those volumes, a DSCB-not-found condition occurs. DFSMShsm is then asked to determine if that data set name is migrated and to recall the data set if it is migrated. Because DFSMShsm does not know that this is in the middle of an uncataloged data set process, it recalls the cataloged, migrated data set and passes control back to open processing. The results of this are indeterminable.