SYNCSORT or DFSORT: Part 1

The sort program is used to sort data into a desired sequence based on  business requirement. Also, it can be used to merge many files into one single file.

Features of the SYNCSORT utility

• Copy Function
• Merge function capable of merging up to 32 data sets into one sequence provided those data sets are sorted. Latest versions allow up to 100 data sets to be merged.
• Sort facility which can sort single or concatenated data sets into one data set.
• Conversion of the packed decimal fields to printable formats.
• Designing of reports

It saves time for developer in writing programs for report, filtering data etc.

SYNCSORT invocation 

• The SYNCSORT is invoked through a JCL
• As a sub-program (in COBOL or PL/1) from a main program

THE JCL Description

1. SORTIN DD

Input file can be:

• PDS Member
• PS sequential file
• VSAM( KSDS,ESDS or RRDS)
• Concatenated File but should be of same record length
•  For Record Format (RECFM) being Variable the file of largest record length and block size if input data sets should be used
• DSB information not be supplied for a disk or standard labelled tape file.

For concatenating data sets , the DCB characteristics of the first data set will be used for the entire file. 

Example of  JCL representation (single file)

 //SORTIN     DD     DSN= DATA1.FILE ,DISP=(OLD,KEEP),

UNIT=TAP38,VOL=SER=10699

 Example of  JCL representation (concatenated files)

 //SORTIN     DD      DSN= DATA1.FILE,DISP=(OLD,KEEP),

//                               UNIT=TAP1,VOL=SER=10699,

//                               DCB= (LRECL=200,RECFM=FB,BLKSIZE=7404

//                    DD     DSN= DATA2.FILE,DISP=(OLD,KEEP),

//                               UNIT=TAP1,VOL=SER=10222

//                    DD     DSN= DATA3.FILE,DISP=(OLD,KEEP),

//                               UNIT=TAP1,VOL=SER=10324

                            

SORTINn DD and SORTINnn are used for merging data sets and the files must themselves be sorted and should be of  the same record format. The valid range of 'n'  is  from 0 to 9 whereas for 'nn' it is 00 to 99 .

Example of JCL representation (merge application)          

                   

 //SORTIN00       DD    DSNAME=DATA4.FILE ,VOL=SER=44444,

//                                   DISP=OLD,UNIT=TAP1

//SORTIN01       DD    DSNAME=DATA5.FILE ,VOL=SER=41235,

//                                   DISP=OLD,UNIT=TAP1

//SORTIN02       DD    DSNAME=DATA6.FILE ,VOL=SER=47876,

//                                   DISP=OLD,UNIT=TAP1

……… and so on.

Here there is no need to supply the DCB information.

2. SORTOUT DD

This is the output file for the SORT,MERGE  or COPY functions and can be any one of the input file types .

Example of a JCL representation

//SORTOUT        DD    DSN=MASTER.OUT,

//                                    DISP=(,CATLG,DELETE),

//                                    UNIT=SYSDA,SPACE=(CYL(100,20),RLSE)

                         

The  DCB Parameters expcept BLKSIZE for the output file will be of the sortin file.

For a merge application the DCB parameters will be the one specified to the first SORTINnn file in the JCL stream.

SORTXSUM DD is the output file that will contain the records eliminated or discarded by SUM processing.

              

SORTWKnn DD: Upto 32 data sets (sort work files) can be allocated. Each SORTWKnn must be allocated on a single unit and a single volume.

Example :  //SORTWK02     DD     UNIT=3380,SPACE=CYL,20         

These SORTWKnn are not required for merge, copy application and when the  sort is executing using DYNALLOC option .

                  

SORTOFxx and SORTOFx are the other statements used to define one or more  output files.

Note: FDR(First Data Resource)  has it's own built-in utility known as FDRSORT which overcomes this problem as it dynamically allocates the necessary sort work space by estimating the number of input records into sort. But this utility cannot be used on a PDS and VSAM file.

 3. SYSIN DD   

SYSIN can be given as control card/file or directly sort statements.

An example is shown below :

Example :  //SYSIN  DD *

 SORT FIELDS = (2,4,CH,A)

                     INCLUDE COND = (15,4,CH,EQ,C'5630')

                  /*

                                                      

4. SYSOUT

This data set contains all the sort messages .

The sort messages can be seen in spool by following statement:

//SYSOUT DD SYSOUT=*    (no data set specified ) .

If you need in dataset, give DSN name instead of SYSOUT.The messages are put in the same msgclass as specified in the job card .

Other DD statements includes $ORTPARM , SORTCKPT , SORTMODS , SYSLIN , SYSLMOD, SYSPRINT(Defines the message data set for the link-editing of sort exits).

     Please refer Sync SORT PART 2 and PART 3.