Data Set Collection

A data set collection is the collection of data sets that are intended to be allocated on the same tape volume or set of tape volumes as a result of data set stacking.

Requirement: All data sets in the data set collection must be directed to the same device category, of the following four device categories:

System-managed DASD
System-managed tape
Non-system-managed DASD
Non-system-managed tape

A mixture of these device categories is not allowed for the following reasons:

There would be problems accessing the data sets later. For example, a data set with a data-set-sequence-number of three could be placed as the first data set on the tape if the first two data sets were redirected to DASD.
There could be problems locating data sets later since some types of data sets must be cataloged and others may be uncataloged.

If you have an allocation in a previous job or step that specifies VOL=SER and that is rerouted to SMS DASD, it will be cataloged on a volume other than that specified in the JCL.

For example, in the following statement, you must not specify volume VOL001 in the JCL unless the volume is in a DUMMY storage group if you later want to allocate the data set as OLD:

//DD1 DD DSN=A,VOL=SER=VOL001,DISP=(NEW,CATLG),LABEL=(1,SL)

Otherwise, it will cause the locate for the data set to be bypassed, probably resulting in an OPEN abend since the data set was rerouted and is not on the volume.

As a result, a statement such as the following is not recognized as data set stacking:

//DD1 DD DSN=A,DISP=OLD
//DD2 DD DSN=B,VOL=SER=VOL001,DISP=(NEW,CATLG),LABEL=(2,SL)

This is because the system does not include existing SMS-managed requests in any data set collection.

When you analyze your current tape usage, you might determine that one of the tape data sets in a data collection would be a good candidate to redirect to DASD. The system attempts to ensure that all data sets in a data set collection are directed to the same device category.