Mainframegeek: January 2011

Thursday, January 20, 2011

The IEBCOPY

IEBCOPY is a dataset utility which used to copy or merge members between Partitioned datasets (PDS) or Partitioned datasets Extended (PDSE), fully or partially.

Common uses of IEBCOPY

1. Copy selected or all members of one PDS into another.

2. Copy a Partitioned dataset into a unique sequential format known as “ Unloaded Partitioned dataset” which can be FTP, manipulate as a sequential dataset or written to Tape.

3. To read an unloaded Partitioned dataset or restore into a partitioned dataset.

4.Compress PDS( c ) in place to recover free space.

The ISPF copy options (3.3) use IEBCOPY under the cover. Moving a PDS or PDSE can be easily achieved by IEBCOPY.

Simple IEBCOPY job

//COPYJOB5 JOB SHIBU,REGION=1024K
//         EXEC PGM=IEBCOPY
//SYSPRINT DD SYSOUT=*
//SYSIN    DD DUMMY
//SYSUT1   DD DISP=SHR,DSN=PGMG.COBOL.SOURCE
//SYSUT2   DD DISP=(NEW,KEEP),UNIT=TAPE,DSN=ARCHIVE.SOURCE,
//   VOL=SER=XXXXXX

SYSUT1 is the input dataset and SYSUT2 is the output dataset. The above job unload the SYSUT1 pds (PGMG.COBOL.SOURCE ) and write into TAPE, with the PDS name provided in SYSUT2.

Note: In the above job output dataset name is not same as the input dataset name, however there is no harm in using the same name.

Job to Restore PDS in a volume.

//COPYJOB6 JOB SHIBU,MSGCLASS=X,REGION=1024K
//         EXEC PGM=IEBCOPY
//SYSPRINT DD SYSOUT=*
//SYSIN    DD DUMMY
//SYSUT1   DD DISP=OLD,UNIT=TAPE,DSN=ARCHIVE.SOURCE,
//   VOL=SER=XXXXXX
//SYSUT2 DD DISP=(NEW,CATLG),DSN=PGMG.COBOL.SOURCE,UNIT=3390,
//   SPACE=(TRK,(10,10,20)),VOL=SER=XXXXXX

This Job read unloaded partitioned dataset from TAPE and write the contents into vol XXXXXX, with the name provided in SYSUT1, 10 tracks and 20 Directory blocks.

We can also use the above job to move datasets from one volume into another with smaller changes.

//COPYJOB7 JOB 1,SHIBU,MSGCLASS=X,NOTIFY=&SYSUID
//         EXEC PGM=IEBCOPY
//SYSPRINT DD SYSOUT=*
//SYSIN    DD DUMMY
//SYSUT1   DD DISP=OLD,DSN=PROD.PGMG.C,UNIT=3390,
//   VOL=SER=VOL1
//SYSUT2 DD DISP=(NEW,CATLG),DSN=SYSI.PGMG.C1,UNIT=3390,
//   SPACE=(TRK,(10,10,20)),VOL=SER=NEWVOL

Instead of using Dummy in the sysin we can specify additional selection criteria to IEBCOPY

E.g.

//SYSIN DD *
  COPY OUTDD=SYSUT2,INDD=SYSUT1
  SELECT MEMBER=(PGM1,PGM2)
/*

SELECT statement specifies the members to be processed with the job. If you have used any other ddnames for the datasets other than SYSUT1 or SYSUT2 then you can use OUTDD, INDD parameters to pass the dataset names.

Restoring a partitioned data set from an unloaded copy automatically compresses (recovers lost space) the data set.

Wednesday, January 19, 2011

Line Sequential File support in Enterprise Cobol.

IBM has added the support for the input and output of line sequential files to Enterprise COBOL. This allows a Cobol program running in mvs to use data stored in zOS/ Unix/HFS file (HFS – Hierarchical File System, is the is the system used to support z/OS UNIX)

Line sequential Files contain only characters(DISPLAY) data and instead of having pre- recorded boundaries, records are delimited by newline characters (x’15’).

Things to consider while writing a COBOL-Line Sequential File program

Select statement

Specify ORGANIZATION IS LINE SEQUENTAL

You cannot open a line sequential file in I-O mode and you cannot use REWRITE statement.

On READ, when a newline character is encountered, it will be replaced with the enough space to fill the record as described after FD.

E.g. If the record length in the line sequential file is 10 and in the FD you given the record length as 200 then after READ, FD-variable will contain 10 characters from Line sequential file followed by 190 blank spaces.

On WRITE trailing spaces are removed and replaced by single newline character .

Things to consider while writing the JCL.

The ASSIGN TO clause in select statement names an external name

At run time, this external name can have a related DD statement that contains

PATH=’file_path’,FILEDATA=TEXT

Sample Program

       Identification Division.
       Program-Id. unixfile.
       Author    . Shibu thannikkunnath.

       Environment Division.
       Input-Output Section.
       FIle-Control.
           Select unixfle assign to indd
           Organization is line sequential
           File status is unixfile-status.


           Select zfsfile assign to zfdd
           File status is zfsfile-status.

       Data division.
       File section.
       FD unixfle.
       01 fs-unix-rec                 pic x(80).

       FD zfsfile.
       01 fs-zfs-rec                  pic x(80).

       Working-storage section.
       01 unixfile-status             pic x(2).
       01 zfsfile-status              pic x(2).

       Eject.

       Procedure division.
       A00100-MAIN.
           Move spaces                 to unixfile-status.
           Perform B00200-WRITE-FILE.
           perform C00300-READ-FILE.
           stop run.

       C00300-READ-FILE.
           Open input unixfle.
           Display 'Read-Module started'.
           If unixfile-status not equal to 00 then
                   Display 'UNIXFILE-STATUS: ' unixfile-status
                   Display 'File is not yet ready, Program terminating'
                   Move 8              to return-code
                   close unixfle
                   stop run
           else
                   Perform until unixfile-status = 'en'
                           read unixfle
                           at end move 'en'
                                       to unixfile-status
                           not at end Display fs-unix-rec
                           end-read
                   end-perform
                   close unixfle
           end-if.
           exit.

       B00200-WRITE-FILE.
           Open output unixfle.
           Open input zfsfile.
           Display 'Write-Module Started'.
           If unixfile-status not = 00 or zfsfile-status not = 00 then
                   Display 'UNIXFILE-STATUS: ' unixfile-status
                   Display 'ZFSFILE-STATUS : ' zfsfile-status
                   Display 'File is not yet ready, Program terminating'
                   Move 8              to return-code
                   close unixfle
                   close zfsfile
                   stop run
           else
                   Perform until zfsfile-status = 'en'
                           read zfsfile
                           at end move 'en'
                                       to zfsfile-status
                           not at end
                               Move fs-zfs-rec
                                       to fs-unix-rec
                               Write fs-unix-rec
                           end-read
                   end-perform
                   Display 'UNIXFILE-STATUS: ' unixfile-status
                   Display 'ZFSFILE-STATUS : ' zfsfile-status
                   close unixfle
                   close zfsfile
           end-if.
           exit.

Sample JCL to execute Program

***************************** Top of Data ******************************
//R0318BEJ   JOB 'SHIBU THANNIKKUNNATH',REGION=0M,NOTIFY=&SYSUID
//EXECPGM     EXEC PGM=UNIXPGM
//INDD        DD   PATH='/u/workds/helloworld.c',
//        FILEDATA=TEXT
//ZFDD        DD   DSN=TSHRCI.PGMG.C(HLLOWRLD),DISP=SHR
//STEPLIB     DD   DSN=TSHRCI.LOAD.LIB,DISP=SHR
//SYSIN        DD   *
/*
//SYSOUT      DD   SYSOUT=*
//
**************************** Bottom of Data ****************************

Mainframegeek

Pages

Thursday, January 27, 2011

z/OS an z/OS.e