Data set requirements

This section discusses the data set requirements for File-AID/Batch, including access method rules, load module copying rules, and input and output data set requirements.

Access method rules

This section discusses rules that are required for accessing input and output data sets. The following access methods are covered:

QSAM
BDAM
VSAM
BPAM
z/OS UNIX zFS

QSAM Access Rules

File-AID accepts all types of QSAM (sequential) data sets as input or output. Keyed QSAM data sets cannot be processed with their keys. QSAM data sets with unlike characteristics can be concatenated.

When processing concatenated data sets, File-AID reports the number of records processed, volume serial number, and dataset name of each concatenated data set. When processing multi-volume data sets, File-AID reports the number of records processed and volume serial number of each volume.

BDAM Access Rules

File-AID accepts all input BDAM formats except formats that are spanned or created with track overflow.

When copying to a BDAM file, and the number of input records is less than the number of available output slots, File-AID does not write null (binary zeros) records to the unused slots.

File-AID uses the BSAM access method to read BDAM data sets.

Keyed BDAM output data sets should be written in ascending key sequence. The key is expected to begin in input record location one (1), followed by the data.

VSAM Access Rules

File-AID accepts four VSAM formats (KSDS, ESDS, RRDS, LDS) containing fixed or variable-length records as input or output.

VSAM data sets are always considered fixed length and contain no RDW. When File-AID copies variable-length VSAM data sets to variable-length sequential output data sets, RDWs are automatically added to the output records. When the output data set is another VSAM data set, File-AID copies the input record length to the output VSAM record length.

When File-AID creates VSAM data sets, they are initially opened with the reset option (load mode). If an error is returned from the open macro, the reset option is turned off and another open macro is tried. If an error is still found, the return code is reported and processing of the data set ceases.

Data sets not opened with the reset option become data set extensions for ESDS or RRDS data sets, and inserts or extensions for KSDS data sets. File-AID provides messages that tell you how the output data sets are opened.

File-AID creates VSAM control blocks during execution with the generate-control-block (GENCB) macro. These control blocks provide transparency between operating systems, and the ability to accept access method changes, which may be implemented by IBM.

One set of control blocks is used for input VSAM data sets and another set is used for output VSAM data sets, except outputs created with the WRITE parameter.

Currently, File-AID/Batch does not support the IBM AMP parameter for defining buffer space for VSAM data sets. It also does not support using the UPDATE function when processing LINEAR data sets (LDS).

BPAM Access Rules

File-AID accepts all PDS record and block formats. It processes concatenated input PDSs with or without a MEMBER parameter. If the MEMBER parameter is used, File-AID processes only the specified members. If the MEMBER parameter is used, File-AID processes only the members that match the specified mask.

You can also select members by using the MBRNAME parameter to identify a range of PDS members by member name.

PDSs with standard ISPF statistics can be processed by testing the statistics with the range parameters: CREATED, CHANGED, and USERID.

When no member filter parameters are specified, File-AID processes all members of all specified concatenated input PDSs.

When processing a PDS with a printing function, File-AID starts a new page and prints the name of each selected member. If you specify PARM=TSO on the EXEC statement, File-AID does not start a new page for each new member name. (For more information, see TSO Execution Parameters).

z/OS UNIX zFS Access Rules

File-AID accepts z/OS UNIX zFS files that are delimited by <newline> characters and treats the records in these files as variable length. z/OS UNIX zFS files that are not delimited by <newline> characters can be treated as fixed length by means of the JCL parameter "LRECL". The valid range is 1 to 32756.

Currently File-AID does not support the concatenation of z/OS UNIX zFS files.

Refer to an IBM JCL Reference Manual for the use of the PATH, PATHDISP, PATHOPTS, and PATHMODE parameters.

Note

The UPDATE function and the BACK modifier cannot be used with z/OS UNIX zFS files.

Load module copying rules

The following rules apply when using File-AID/Batch to copy load modules from one load library to another:

Both the output and the input data set must be partitioned and must have an undefined record format (RECFM=U).
- If the input is a PDSE load library, the output can be either a PDS or PDSE load library. If the output is PDSE load library then the input must also be a PDSE load library.
- PDSE load libraries cannot be concatenated.
- The following are not supported for PDSE load libraries:
  - AUDIT
  - COPY with IF
  - COPY with REPL
  - COPYALL
  - COPYMEM
  - DROP
  - SPACE
  - UPDATE/UPDATEALL
  - USER
Reblocking is not supported. The block size of the output data set must be equal to or larger than the block size of the input data set. The logical record length is ignored.
Copying load modules created for planned overlay, scatter-loaded, or note listed modules is not supported.
Copying aliases is supported. If an alias is selected along with the base member to which the alias refers, the base member will be copied and the alias will be stored as an alias of the base. If an alias is selected without the base member, then the alias will be copied as a unique copy of the base member and will not appear as an alias in the output data set.

Input data set requirements

File-AID can process 1 to 100 input data sets during an execution with no limitation on type or mixture of data set organizations.

Blocked and Unblocked Data sets

File-AID processes blocked or unblocked data sets using the data-control-block (DCB) information on the input data set. To override the input data set’s DCB settings, enter the desired record format (RECFM), record length (LRECL), and block size (BLKSIZE) in the DCB field of the input DD statement.

Variable-Blocked-Spanned Data sets

File-AID automatically processes variable-blocked-spanned (VBS) data sets in their completed format. If VBS data sets must be processed as segments, code the RECFM on the input DD statement (//DDxx) as RECFM=VB. File-AID does not process VBS BDAM data sets in the completed format.

Proprietary Source Library Data sets

File-AID supports several proprietary source library formats as primary input (DD01) including:

CA Panvalet
CA Librarian.

Note

File-AID enables you to COPY from (but not to) a CA Panvalet or CA Librarian data set. The CA Panvalet or CA Librarian data set cannot be the destination data set (DDxxO) in a COPY statement.
File-AID supports only 80-character CA Panvalet or CA Librarian members. When greater than 80 characters, File-AID will truncate records greater than 80 characters and issue an error message.

Some additional installation steps may be required to activate support. Refer to File-AID Installation and Configuration Guide.

Tape Data sets

File-AID processes tape input using the data set information on the tape label. If the tape is unlabeled or non-standard labeled, File-AID uses the DCB defaults RECFM=U and BLKSIZE=32767 unless you override. If needed, use JCL parameters to override label information. File-AID supports LRECL up to 32,760 and BLKSIZE up to the maximum for the device, such as 256K for 3590 tape unit.

Variable-Length Record Data sets

File-AID can copy variable-length data sets to fixed-length data sets. Ensure that data is not truncated in the operation. The record descriptor word (RDW) parameter specifies the handling of the RDW word for printing and processing.

Unit Affinity Statement

When File-AID processes more than one data set, and each resides on the same physical storage unit, use the unit affinity option to minimize unit allocation time.

File-AID works with each data set separately; it always closes the current data set before opening the next one. Therefore, multiple input or output data sets can logically occupy the same physical units to reduce excessive unit allocation. To do this, code the unit affinity statement in the input DD statements as shown in the following example:

//DD01 DD DSN=NAME1,UNIT=TAPE
//DD02 DD DSN=NAME2,UNIT=AFF=DD01

Important

If you intend to write the USER function output to tape, read Examples (Example 35 and 36).

Output data set requirements

A maximum of 100 output data sets of any data set organization or access method can be created during a File-AID execution. Create output data sets with the COPY, DROP, REFORMAT, or USER functions. When you create data sets with the COPY, DROP or REFORMAT functions, records are written to the corresponding output //DDxxO DD data set. When you create data sets with the USER function, the WRITE=anyname parameter directs output to a //anyname DD. See JCL Required for Execution.

You can write output data sets to tape, disks, cards, or printers. File-AID copies basic DCB information from the input data set when no DCB information is specified for the output data set. If needed, use JCL parameters to specify new label information. File-AID supports LRECL up to 32,760 and BLKSIZE up to the maximum for the device, such as 256K for 3590 tape unit.