Automating Data Allocation Using Data Classes

TSO data set naming conventions are based on the high-level qualifier, typically the TSO user ID. You can use the data set's LLQ to identify the type of data. Some examples of LLQs are:

PLI: for PL/1 source programs
CLIST: for CLIST libraries
CNTL: for JCL libraries
TEXT: for TEXT data
LOAD: for load libraries
LIST: for listing data sets

The starter set assumes that you have a data set naming convention based on LLQ. Sample data classes are included in the starter set to help you define the following:

VSAM data sets based on the RECORG parameter
With these data classes, users can create key-sequenced, relative record, entry-sequenced, or linear VSAM data sets using batch JCL run in the background, or TSO ALLOCATE commands run in the foreground. These data classes create VSAM data sets with a primary allocation of 400 KB. The user supplies the information about key offset and length for VSAM key-sequenced data sets. Any performance-related options are the user's responsibility to provide.

Table 1 shows the attributes for these sample VSAM classes.
Simple physical sequential data sets including test data sets and output listings.
You can create 80-byte, fixed-block data sets by using the class DATAF. The primary space requested is 400KB.

Data sets having variable-blocked records with a average record size of 255 bytes are defined using DATAV. Based on the primary space request, 1.275 MB are allocated.

Listing data sets having a primary space of 90 MB are allocated using the LISTING data class.

Table 2 shows the attributes for these physical sequential data classes.
Load and source libraries in both PDS and PDSE format
The data class, LOADLIB, is reserved for load libraries that you intend to allocate as partitioned data sets. SRCFLIB and SRCVLIB are data classes that allocate PDSEs, based on the value of the DATA SET NAME TYPE attribute.

Table 3 shows the attributes for these model libraries.