DD Statement

It is the Data Definition(DD) statement for the JCL which provides information about the dataset and to specify the input, output resources needed for that JOB. A maximum of 3273 DD statements can be coded per job STEP.

//DDNAME    DD     ,

The DD statement is used to

• Describe a dataset and to specify the input and output resources needed for the dataset.
• Map a logical file used by the processing program to a physical dataset.

Rules to follow for DD statement:

• Each ddname should be unique within the job step.
• The ddname must begin in column 3.
• The ddname is 1 through 8 alphanumeric or national ($,#,@) characters.
• The ddname must be followed by atleast one blank.
• The ddname is omitted if it defines a dataset that is to be concatenated to the dataset of the preceding DD statement.

DDName:

It is a mandatory field. If omitted, the dataset is concatenated with the one defined in the previous DD statement.

//PROGPUB JOB NOTIFY=&SYSUID,MSGCLASS=A,MSGLEVEL=2
//STEP1   EXEC PGM=PROGRAM1
//INPUT   DD DSN=PROGPUB.FLATFILE.INPUT,DISP=SHR
//OUTPUT  DD DSN=PROGPUB.FLATFILE.OUTPUT,DISP=(NEW,CATLG,DELETE),
//            RECFM=FB,LRECL=80,SPACE=(TRK,(2,5)) 

Positional Parameters:

The most widely used positional parameters are explained here

*(asterisk):

Use of the *(asterisk) parameter to begin an instream data set. It is also used to pass data to the program like the PARM parameter in EXEC statement. The '/*' indicated the end of instream data. When the instream data is passed using '*', the preceding lines should not contain any '//' in the column 1 and 2.

//PROGPUB JOB NOTIFY=&SYSUID,MSGCLASS=A,MSGLEVEL=2
//STEP1   EXEC PGM=PROGRAM1
//DD1      DD *
INSTREAM DATA
/* 

When using an utility program like SORT, the in-stream data is used to pass the SORT CARD parameters.

//PROGPUB JOB NOTIFY=&SYSUID,MSGCLASS=A,MSGLEVEL=2
//STEP1   EXEC PGM=DFSORT
//DD1     DD *
  SORTFIELDS=(CONDITION)
/*

DATA:

DATA parameter is functionally same as '*' operand except the instream data can contain '//' in column 1 and 2.

//PROGPUB JOB NOTIFY=&SYSUID,MSGCLASS=A,MSGLEVEL=2
//STEP1   EXEC PGM=PROGRAM1
//DD      DD DATA
//INSTREAM DATA
/* 

Incase if the instream data has the value '/*' which is also an indicator to define the end of instream data, we can define a custom delimiter as shown in the example

//PROGPUB JOB NOTIFY=&SYSUID,MSGCLASS=A,MSGLEVEL=2
//STEP1   EXEC PGM=PROG1
//DD1     DD DATA,DLM='$$'
/*INSTREAM DATA
$$

DUMMY or NULLFILE:

The DUMMY or NULLFILE keyword is used to assign a dummy status to a dataset so any input-output operation on the dummy dataset is bypassed.

//PROGPUB JOB NOTIFY=&SYSUID,MSGCLASS=A,MSGLEVEL=2
//STEP1   EXEC PGM=PROGRAM1
//DD1     DD DUMMY

Keyword Parameters:

There are wide range of keyword parameter. The most common are explained below,

DSN:

This parameter is used to define the permanent or temporary dataset name. Dataset name is maximum 44 characters long with a period after 8 characters or less. Maximum length is inclusive of periods. Temporary datasets are used a work area for the duration of the job.
Syntax is,

DSN=DATASET NAME

//PROGPUB JOB  NOTIFY=&SYSUID,MSGCLASS=A,MSGLEVEL=2
//STEP1   EXEC PGM=PROGRAM1
//DD1     DD   DSN=PROGPUB.FLATFILE.INPUT

To define a temporary dataset, specify a name starting with && followed by 1 to 6 characters or without specifying DSN.

DSN=&&TEMP

//PROGPUB JOB  NOTIFY=&SYSUID,MSGCLASS=A,MSGLEVEL=2
//STEP1   EXEC PGM=PROGRAM1
//DD1     DD   DSN=&&TEMPDT

DISP:

This parameter is used to specify the status of the dataset and to specify what to do with the dataset after termination (normal or abnormal) of the step or the job. The disposition status specifies the existence of a dataset at the start of the STEP. Syntax is

DISP=(Status,Normal-End,Abnormal-End)

Status:

Normal-End:

This disposition parameter specifies what to do with the dataset after a normal termination. It takes one of the values,
CATLG, UNCATLG, PASS, KEEP, DELETE.

Abnormal-End:

This disposition parameter specifies what to do with the dataset after a normal termination. It takes one of the values,
CATLG, UNCATLG, KEEP, DELETE.

To create a new member in an existing PDS, code DISP=OLD or SHR, not NEW, since DISP refers to the partitioned dataset as a whole not to the individual member. The PDS directory is updated about the new member.
For VSAM datasets, the disposition should be coded as DISP=OLD or DISP=SHR.

UNIT:

This parameter should be specified for new and existing uncatalogued datasets. Syntax is

UNIT=(Hardware Address or Devce model number or symbolic group name or virtual I/O)

//PROGPUB JOB  NOTIFY=&SYSUID,MSGCLASS=AMSGLEVEL=2
//STEP1   EXEC PGM=PROGRAM1
//DD1     DD   DSN=PROGPUB.FLATFILE.NEW,DISP=(NEW,CATLG,DELETE)
//              UNIT=SYSDA

VOL:

This parameter goes along with the UNIT parameter. The VOL parameter specifies the name and number of the volume of the UNIT specified. Volume is the name of a specific UNIT. One unit can have multiple volumes.

//PROGPUB JOB  NOTIFY=&SYSUID,MSGCLASS=A,MSGLEVEL=2
//STEP1   EXEC PGM=PROGRAM1
//DD1     DD   DSN=PROGPUB.FLATFILE.NEW,DISP=(NEW,CATLG,DELETE)
//              UNIT=(3390,2),VOL=SER=(VOL01,VOL02)

SYSOUT:

This parameter is specified to redirect the data to output device based on the SYSOUT class specified. Syntax is

SYSOUT=CLASS

//PROGPUB JOB  NOTIFY=&SYSUID,MSGCLASS=A,MSGLEVEL=2
//STEP1   EXEC PGM=PROGRAM1
//DD1     DD   DSN=PROGPUB.FLATFILE.NEW,DISP=(NEW,CATLG,DELETE)
//              UNIT=SYSDA
//DD2     DD   SYSOUT=*

SPACE:

This parameter is used to specify the amount of space required by the new dataset in DASD. A smallest unit of space in mainframe is one track and group of tracks are called Cylinders. One or more contiguous tracks or cylinders allocated to a dataset is called EXTENT Syntax is

SPACE=(UNIT OF SPACE,(PRIMARY,SECONDARY,DIRECTORY)RLSE,CONTIG,ROUND)

//*JCL with SPACE=(TRK,(2,5))
//PROGPUB JOB NOTIFY=&SYSUID,MSGCLASS=AMSGLEVEL=2
//STEP1   EXEC PGM=PROGRAM1
//INPUT   DD DSN=PROGPUB.FLATFILE.INPUT,DISP=SHR
//OUTPUT  DD DSN=PROGPUB.FLATFILE.OUTPUT,DISP=(NEW,CATLG,DELETE),
//            RECFM=FB,LRECL=80,SPACE=(TRK,(2,5))

For this example, the space allocation can be calculated as
2 primary tracks + (5*15) secondary tracks = 77 tracks

//* Space allocation requested in blocksize
//OUTPUT  DD DSN=PROGPUB.FLATFILE.OUTPUT,DISP=(NEW,CATLG,DELETE),
//            RECFM=FB,LRECL=80,SPACE=(800,(2,5))

For this example, the space allocation can be calculated as
2 blocks each holding 800 bytes. So, total primary blocks 1600 bytes.
5*15*800 bytes are allocated for secondary storage.

Examples
SPACE=(CYL,10)
SPACE=(1024,75)
SPACE=(TRK,(50,10,5))
SPACE=(TRK,(100),RLSE)
SPACE=(CYL,(10,,10),,CONTIG)

DCB:

This parameter is called Data Control Block and It defines the dataset characteristics. Syntax is

DCB=(RECFM,LRECL,BLKSIZE,DSORG)

RECFM:

The RECFM parameter specifies the record format of the dataset. It takes the value as listed below

LRECL:

This parameter defines the length of the logical records. For a fixed length records, the dataset contains all of the same length and hence LRECL specifies the actual length of the data.
For variable length records, the length of the records vary from an average to a maximum. So, the LRECL = longest data-bytes length +4 bytes. The 4 bytes are called Record Descriptor Word (RDW) is pre-penned to each record and indicates the actual length of the record in its first 2 bytes.
For Undefined format, the parameter is coded as LRECL=0.

BLKSIZE:

It defines the size of a block of record. Blocking of records improves processing speed. For fixed length blocked record the value for blocksize is specified in multiples of LRECL. For variable length blocked record, the value for blocksize is specified in multiples of LRECL plus 4 bytes.

//* Space allocation requested in blocksize
//OUTPUT  DD DSN=PROGPUB.FLATFILE.OUTPUT,DISP=(NEW,CATLG,DELETE),
//             DCB=(RECFM=FB,LRECL=80,BLKSIZE=8000,SPACE=(800,(2,5))

DSORG:

This parameter specifies the organization of the data set as physical sequential (PS), partitioned (PO), or direct (DA).