Skip to main content

Introduction​

IDCAMS (Integrated Data Set Control Access Method Services) is a utility provided by IBM to manage datasets and catalogs, especially for VSAM and GDG datasets. It allows users to create, delete, define, rename, back up, and list information about both VSAM and non-VSAM datasets. IDCAMS commands are typically used inside JCL or TSO batch jobs to perform catalog-related operations efficiently.

IDCAMS plays a key role in VSAM dataset lifecycle management, GDG operations and is essential for both system programmers and application developers working on IBM Mainframes. Here are some common Knowledge information around IDCAMS.

Some of the most common IDCAMS commands include:

Exercise - Create IDCAMS SYSIN​

Write an IDCAMS SYSIN for this logic for below requirements.

  • Delete a dataset and only define it if the delete command is successful.
  • You need to delete a dataset if it exists, and then define it a new. However, if the dataset doesn't exist, you must skip the delete command and move directly to the define step.
  • Checks if a dataset named SHRDVXX.DATASET exists. If it exists, print its contents; if it doesn’t exist, define it.
  • Define a VSAM KSDS cluster with embedded INDEX and DATA components. Include space allocation for both.
  • Delete a dataset if it exists. If not to terminate further execution of IDCAMS.
  • Define an ESDS dataset with the following properties:
    • Dataset name: SHRDVXX.ESDS
    • Record size: 100 100
    • Space allocation: 10 5
    • Control Interval Size: 2048
  • Define a non-VSAM dataset (a simple sequential dataset) named TEXT.FILE. Make sure to use appropriate options.

Exercise - Fix Invalid IDCAMS sysin​

SET MAXCC = 16
DELETE MY.DATASET
DEFINE CLUSTER(NAME(MY.DATASET) ...)
DEFINE CLUSTER(NAME(MY.DATASET) - 
               KEYS(10 0) -
               RECORDSIZE(80 80)
               TRACKS(5 5))
IF LASTCC = 0 THEN
   DELETE MY.DATASET
END IF
DEFINE CLUSTER(NAME(USER.KSDS) - 
               KEYS(10 0) RECORDSIZE(80 80) -
               TRACKS(5 5))

Exercise: Invalid IDCAMS Syntax​

Steps:

  1. Use a DEFINE command with a typo (e.g., misspelling RECORDSIZE to something like RECORSIZE).
  2. Submit the JCL.

Questions:

  • What message do you receive in SYSOUT?
  • Does IDCAMS stop processing further commands?

Exercise: IDCAMS REPRO​

Steps:

  • Create a KSDS.
  • Try to use REPRO to copy the same file into it twice:
  • Then run the same REPRO again.

Questions:

  • What happens the second time? Do duplicates get inserted?
  • What behavior do you observe with respect to keys?
  • How can you prevent duplicate insertions?

Exercise: DELETE vs DEFINE on VSAM​

Steps:

  • DELETE the KSDS cluster only:
//DELKSDS  EXEC PGM=IDCAMS
//SYSIN    DD *
   DELETE YOUR.USER.KSDS CLUSTER
   SET MAXCC = 0
/*
  • LISTCAT on YOUR.USER.KSDS.
  • Attempt to DEFINE the same cluster again with different RECORDSIZE.

Questions:

  1. After DELETE CLUSTER, does LISTCAT still show any entry?
  2. When you run DEFINE again, do you get an error or does it succeed?
  3. What remnants (catalog entries or control blocks) remain on disk?

Exercise: Difference between Define Cluster​

  • Here are the 2 JCLs to allocate KSDS.
//DEFKSDS  EXEC PGM=IDCAMS
//SYSPRINT DD SYSOUT=*
//SYSIN    DD *
   DEFINE CLUSTER (
      NAME(YOUR.USER.KSDS) 
      INDEXED 
      RECORDSIZE(80,80) 
      KEYS(8,0) 
      CISZ(4096) 
      VOLUMES(YOURVOL)
   )
/*
//CREATEK EXEC PGM=IDCAMS
//SYSIN   DD *
   DEFINE CLUSTER (NAME(YOUR.USER.KSDS1) -
                  INDEXED -
                  KEYS(8 0) -
                  RECORDSIZE(80 80) -
                  TRACKS(1 1)) -
      DATA(NAME(YOUR.USER.KSDS1.D)) -
      INDEX(NAME(YOUR.USER.KSDS1.I))
/*

Questions:

  1. Run LISTCAT job to find out what is the difference details you see for YOUR.USER.KSDS1 and YOUR.USER.KSDS?
  2. What is the CISZ for both the datasets?
  3. Modify the DEFINE CLUSTER command for YOUR.USER.KSDS1 to omit the INDEX portion and observe the result.

Exercise: Define, List, and Print a VSAM KSDS​

Steps:

  • Run this JCL to DEFINE a small KSDS
//DEFKSDS  EXEC PGM=IDCAMS
//SYSPRINT DD SYSOUT=*
//SYSIN    DD *
   DEFINE CLUSTER (
      NAME(YOUR.USER.KSDS) 
      INDEXED 
      RECORDSIZE(80,80) 
      KEYS(8,0) 
      CISZ(4096) 
      VOLUMES(YOURVOL)
   )
/*
  • LIST its catalog entry
//LISTCAT  EXEC PGM=IDCAMS
//SYSIN    DD *
   LISTCAT ENT('YOUR.USER.KSDS') ALL
/*
  • Attempt to PRINT from it:
//PRTKSDS  EXEC PGM=IDCAMS
//INFILE   DD DSN=YOUR.USER.KSDS,DISP=SHR
//SYSPRINT DD SYSOUT=*
//SYSIN    DD *
   PRINT INFILE(YOUR.USER.KSDS) COUNT(5)
/*

Questions:

  1. What RECORDSIZE, KEY length, and CISZ values does LISTCAT report?
  2. Does the PRINT step succeed on an empty KSDS? If not, what message appears?
  3. What happens if you skip the DEFINE and rerun the PRINT step?

Exercise: Load Records with REPRO​

Steps:

  • Create a 5-record sequential dataset (80 bytes each).
  • REPRO into the KSDS:
//LOAD     EXEC PGM=IDCAMS
//SYSIN    DD *
   REPRO INFILE(INPUT) OUTFILE(YOUR.USER.KSDS)
/*
  • REPRO the first 5 records again.

Questions:

  1. After REPRO, what does LISTCAT show for record count?
  2. Does PRINT now display your test records?
  3. If your INPUT file contains duplicate keys, what return code does REPRO return?

Exercise: GDG Referencing​

Steps:

  • Create a GDG Base. Ensure your GDG base is empty.
  • Create the new Generation using (+1) reference.
  • Execute below job with necessary JOB Statement
//STEP2  EXEC PGM=IEFBR14
//DD1    DD DSN=YOUR.GDG.BASE(0),DISP=OLD
//DD2    DD DSN=YOUR.GDG.BASE.G0001V00,DISP=OLD

Questions:

  • Which dataset does DD1 read?
  • Which dataset does DD2 read?
  • If you change DD2 to YOUR.GDG.BASE(+1), what happens?

Exercise: Skipping Generation Numbers​

Steps:

  • Create a GDG Base with LIMIT(3) EMPTY SCRATCH. Ensure your GDG base is empty.
  • Allocate 3 dataset with following generation refernce (+1), (+10), (+11).
  • Allocate One more dataset in new job using (+1).

Questions:

  • After first job, list the cataloged generations.
  • What does (0) point to? What does (-1) point to?
  • After SKIP2, which generations remain?

Exercise: GDG DISP Combinations​

Steps:

  • Create a job and execute with 2 step. In step 1 allocate the generation using (+1) and DISP=(NEW,DELETE,DELETE). In step 2 allocate the generation using (+1) and DISP=(NEW,CATLG,DELETE).
  • Create a job to read the generation using (0) reference and DISP=OLD.

Questions:

  1. Which generation (if any) is cataloged after first job?
  2. What happens when READ uses (0)?
  3. If you swap step 1 and step of Job 1 (i.e. catalog first, then delete), how does the result change?

Exercise: GDG Manual Catalog Deletion​

Steps:

  • Create three generations properly cataloged: (+1), (+2), (+3).
  • Manually delete the second generation.
  • In a new job read following job.
//PROBE  EXEC PGM=IEFBR14
//L0     DD DSN=YOUR.GDG.BASE(0),DISP=OLD
//L1     DD DSN=YOUR.GDG.BASE(-1),DISP=OLD
//L2     DD DSN=YOUR.GDG.BASE(-2),DISP=OLD

Questions:

  1. What does L0 read?
  2. What does L1 read?
  3. What does L2 do or read?