Partitioned Concatenation

To process partitioned concatenated data sets, use a DCB that has DSORG=PO. When PDSEs are concatenated, the system treats the group as a single data set. A partitioned concatenation can contain a mixture of PDSs, PDSEs, and UNIX directories. Each PDSE is treated as if it had one extent, although it might have multiple extents. You can use partitioned concatenation only when the DCB is open for input.

There is a limit to how many DD statements are allowed in a partitioned concatenation. Add together the number of PDS extents, the number of PDSEs, and the number of UNIX directories in the concatenation. If the sum exceeds 255, then the OPEN macro results in an ABEND 013-E4. For example, you can concatenate 15 PDSs of 16 extents each with 8 PDSEs and 7 UNIX directories ((15 x 16) + 8 + 7 = 255 extents).

Concatenated PDSEs are always treated as having like attributes, except for block size. The concatenation uses only the attributes of the first data set, except for the block size. BPAM OPEN uses the largest block size among the concatenated data sets. For concatenated fixed-format data sets (blocked or unblocked), the logical record length for each data set must be equal.

Process a concatenation of PDSEs in the same way that you process a single PDSE, except that you must use the FIND macro to begin processing a member. You cannot use the POINT (or NOTE) macro until after you issue the FIND macro for the appropriate member. If two members of different data sets in the concatenation have the same name, the FIND macro determines the address of the first one in the concatenation. You would not be able to process the second data set in the concatenation. The BLDL macro provides the concatenation number of the data set to which the member belongs in the K field of the BLDL list. (See BLDL—Construct a directory entry list.)