VSAM

Overview
 VSAM (Virtual Storage Access Method )

VSAM is an integral part of MVS. At the end of this course,

you, the user, will know VSAM in detail.
 know the different types of VSAM data sets
 be able to create, delete and alter VSAM data sets, with
indexes and alternate indexes
 know the organization of VSAM data sets
 Access Methods

 An access method defines the technique by which data is

stored and retrieved. The different types of dataset
organizations in MVS are:
 Physical Sequential
 Partitioned
 Indexed Sequential
 Direct
 VSAM
 What is VSAM?

 VSAM is Virtual Storage Access Method

 It is a method used to move data between Disk and Main
Storage
 VSAM operates in Virtual Environment
VSAM

 VSAM acts as interface between

 Operating System and Application Program

V
Request S Reply Operating
Application
A System
M
 VSAM

 Interface between Main Storage and Disk

V
S
Main Storage DATA A DATA Disk
M
 File access methods

 Data (Records) is retrieved

 Sequential (Reading from beginning to end)
 Random (Records are read by the value in the key)
 Direct (Records are read based on their physical
location/address on disk)
 VSAM provides all these methods
 One access method supporting all types of data retrieval
 Traditional access methods

 QSAM (Queried Sequential Access Method)

 BSAM (Basic Sequential Access Method)
 for ‘flat’ files
 ISAM (Index Sequential Access Method)
 for Index files
 BDAM (Basic Direct Access Method)
 for direct access files
 VSAM Dataset Types

Entry Sequenced Dataset (ESDS)

 ESDS contains records in the order in which they are entered. Records
are added to the end of the data set, and can be accessed sequentially.
Key Sequenced Dataset (KSDS)
 KSDS contains records in ascending collating sequence. Records can
be accessed by a field called a key, or by a relative byte address (RBA -
relative position of the record from the beginning of the dataset), or
sequentially.
Linear Dataset (LDS)
 LDS contains data that has no record boundaries. The manipulation of
the data is completely controlled by the user.
Relative Record Dataset (RRDS)
 RRDS contains records in the order of relative record number. These
records can be accessed only by this number.
 VSAM data set organization

 VSAM Data Set can contain three major components

 CLUSTER (Catalog entry)
 INDEX
 DATA (Actual data)
 Data Set is referred by cluster name in JCL

INDEX
CLUSTER
DATA
 VSAM internals

 CONTROL INTERVAL (CI)

 VSAM stores Data and Index in Control Intervals (CI)
 CI is similar to ‘Block’

CI

Record Record .......

 CONTROL INTERVAL

 CI is the unit of retrieval between DASD and I/O Buffer (Virtual

Storage)

DASD I/O Buffer

CI

R1 R2 R3 R1 R2 R3
 Control Interval

 CI contains
 Records (or DATA)
 Free space (Optional)
 Control Information Definition Field (CIDF)
 Record Definition field (RDF)
 Building blocks of a VSAM Dataset

Logical Record
 Logical records of VSAM data sets are stored in a different manner from logical
records in non-VSAM data sets. VSAM stores records in control intervals.
Control Interval (CI)
A control interval consists of the following:
 Logical records (LR)
 Free space (FS)
 Control information fields

LR LR LR LR..FS....FS...RDF CIDF

 Free Space is the space reserved within the CI for inserting new records in a KSDS or
for lengthening the existing records.
 CONTROL INTERVAL

 CIDF & RDF are VSAM control functions

 Used by VSAM to access data

R R C
R1 R2 Free D I
R3 D
Space F D
F
F

3 bytes, Length of Record

How many successive
records have same length 4 bytes long, One per CI
Indicates Free space,
where and how much
 CONTROL AREA

Control Area
 CIs are grouped into CA
 The control intervals in a VSAM data set are grouped together
into Control Areas.
 A VSAM data set is actually composed of one or more control
areas. The maximum size of a control area is a cylinder, and
the minimum size is one track.

Control Area
(CA)
LR LR LR LR.. FS....FS... RDF CIDF
LR LR LR LR.. FS....FS... RDF CIDF
LR LR LR LR.. FS....FS... RDF CIDF
 Cluster

 VSAM datasets are defined as Highest Index Set

Clusters. 2210 5000

 A Cluster is a combination of the

data component and the index Index Set
component. 1055 1600 FS 5000 FS FS

 The Index component is applicable

only in the case of a KSDS. Sequence Set
 The data component holds data 1010 1055 FS 1350 1600 FS
records.
 The index component holds the
index records to access the required DATA COMPONENT (CA1)
information from the data
component 1300 1305 1310 1350 FS
1400 1410 1415 1600 FS
Data Component (CA2)
FS FS FS FS FS
1001 1002 1005 1010 FS
1020 1022 1030 1055 FS
FS FS FS FS FS
 Control Interval Split

CA-1 (CI-1 to CI-3) - Before split

 When a data set is first loaded,
the key sequence of records and 1001 1002 1005 1010 FS
their physical order are the same.
 However, when records are 1020 1022 1030 1055 FS
inserted, control interval split can
occur, causing the data control
intervals to have a physical order FS FS FS FS
that is different from the key
sequence.
 For example, try to insert '1004' CA-1 (CI-1 to CI-3) - After split
in CI-1 shown below :
 A Control Interval Split occurs
1001 1002 1004 FS
and CI-1 is split exactly into two
half CI's. CI-3, which is a free 1020 1022 1030 1055 FS
control interval is used, and some
logical records of CI-1 are moved
into CI-3. CI-3 is placed after CI- 1005 1010 FS FS
2 and it is not inserted in between
CI-1 and CI-2.
 Control Area Split

 Now, if record ‘1025’ is inserted, there is no free control

interval for a CI split to occur. Hence, a CA (control area)
split occurs. The CA-1 is split into two half control areas;
some of the Control intervals of CA-1 are moved into the
newly created CA (CA-2).
CA-1 (CI-1 to CI-3) - After split

1001 1002 1004 FS

1020 1022 FS FS FS
FS FS FS FS

CA-2 (CI-1 to CI-3)

1025 1030 1055 FS

1005 1010 FS FS
FS FS FS FS
 Types of VSAM data sets

 ESDS Entry Sequenced Data Set

 KSDS Key Sequenced Data Set
 RRDS Relative Record Data Set
 LDS Linear Data Set
 ORGANIZATION

SEQUENTIAL => ESDS

INDEXED => KSDS
RELATIVE => RRDS

ACCESS MODE OPTIONS

ESDS : SEQUENTIAL
KSDS : SEQUENTIAL or RANDOM or DYNAMIC
RRDS: SEQUENTIAL or RANDOM or DYNAMIC
 Access Method Services (AMS)

 Access Method Services is a service program that helps you to

allocate and maintain VSAM data sets and catalogs.
 IDCAMS is the name of the utility program used to perform the
following operations on VSAM data sets.
 Creating a Data set

 Loading a VSAM data set.

 Printing a Data set

 Deleting a Data set

 Error Detection for KSDS Data set

 Creating Generation Data Groups (GDG) etc.

INVOKING ACCESS
METHOD

IDCAMS UTLITY
 Defining a VSAM Cluster

 The Define Cluster command is used to allocate VSAM data sets. The basic information
required for defining a VSAM data set is:
 Name of the cluster.
 Volume(s) on which the data set will be allocated.
 Type of data set (KSDS, ESDS, RRDS or LDS).
 Space needed for the data set.
 For a KSDS, the length of the key and its offset from the beginning of the record.
 Record size, and whether it is fixed or variable in length.
 Control interval size.
 CI and CA Free Space.
 The following keywords have to be used to define the different types of VSAM data sets:
 NONINDEXED for ESDS.
 INDEXED for KSDS.
 NUMBERED for RRDS.
 LINEAR for LDS.
 Sample JCL to create an ESDS cluster

//LEM0UXXA JOB
MSGCLASS=Q,CLASS=Q,NOTIFY=&SYSUID
//DEFKSDS EXEC PGM=IDCAMS NONINDEXED - ESDS
//SYSPRINT DD SYSOUT=*
//SYSIN DD *
DEFINE CLUSTER(NAME(LEM0UXX.TEST.ESDS) -
VOLUMES(LEM001) -
RECORDSIZE(N1,N2)
NONINDEXED -
TRACKS(2,1) - N1=> AVG RECL
RECORDSIZE(50,50) - N2=> MAX RECL
CONTROLINTERVALSIZE(4096) -
FREESPACE(10,20)) -
DATA(NAME(LEM0UXX.TEST.ESDS.DATA))
/*

Freespace(PCT1,PCT2)

PCT1=> PCT FREE SPACE IN each CI

PCT2=> PCT Of unused CI in CA

COBOL AND ESDS

OVERVIEW
 ESDS

 Similar to Sequential File

 Sequenced by the order in which data is entered/loaded
 New Records are added at the end only (chronological order)
 Supports both Fixed and Variable formats
 Contains only CLUSTER & DATA components
 ESDS (Contd...)

 Only sequential access in Batch Cobol Programs

 Random access is supported in on-line applications (CICS)
using Relative Byte Address (RBA)
 Alternate Index is supported in on-line applications (CICS)
 NO primary index
 RBA

 Record location relative to the beginning of the file

(Relative Byte Address)

CI

R1 R2 R3
80 40 60

RBA of R1 is 80
 FILE-CONTROL Paragraph

Format:

SELECT [OPTIONAL] File-name-1 ASSIGN TO AS-Assignment-name-1

SEQUENTIAL
[ ORGANIZATION IS INDEXED
RELATIVE

SEQUENTIAL
[ACCESS MODE IS RANDOM
DYNAMIC

[FILE STATUS IS Data-name-1]

 ESDS

 ORGANIZATION IS SEQUENTIAL
 ACCESS MODE IS SEQUENTIAL

INPUT OUTPUT INPUT- EXTEND

OUTPUT
READ YES - YES -

WRITE - YES YES

DELETE

REWRITE YES
VSAM === ESDS

ORGANIZATION IS SEQUENTIAL.
ACCESS MODE = SEQUENTIAL

OVERVIEW
CREATING ESDS
Writing Data in ESDS
Writing Data in ESDS
Writing Data in ESDS
RUNNING A PROGRAM
View the data from ESDS

REPRO COMMAND
PRINT ESDS EXAMPLE

Overview
PRINT ESDS
PRINT ESDS
PRINT ESDS
KSDS (KEY SEQUENTIAL
DATA SET )
 Sample JCL to create a KSDS cluster
KEYS(N1,N2)

//LEM0UXXA JOB MSGCLASS=Q,CLASS=Q,NOTIFY=&SYSUID N1=> length of the Key(Bytes)

//DEFKSDS EXEC PGM=IDCAMS N2=> is starting byte position

//SYSPRINT DD SYSOUT=* of Key in Record

//SYSIN DD *

DEFINE CLUSTER(NAME(LEM0UXX.TEST.KSDS) - RECORDSIZE(N1,N2)

VOLUMES(LEM001) - N1=> AVG RECL
TRACKS(2,1) -
N2=> MAX RECL
INDEXED -

KEYS(9,0) -
INDEXED - KSDS
RECORDSIZE(50,50) -

CONTROLINTERVALSIZE(4096) -
Freespace(PCT1,PCT2)
FREESPACE(10,20)) -
PCT1=> PCT FREE SPACE IN each CI
DATA(NAME(LEM0UXX.TEST.KSDS.DATA)) -
PCT2=> PCT Of unused CI in CA
INDEX((NAME(LEM0UXX.TEST.KSDS.INDEX))
 SELECT & ASSIGN Syntax

ENVIRONMENT DIVISION.
INP UT- OUT P UTSECT ION.
FILE - CONT ROL.
SELECT [OP T IONAL] FileName
ASSIGN T O FileSpec
ORGANIZAT ION IS INDEXED
SEQUENT IAL
ACCESS MODE IS RANDOM
DYNAMIC
RECORD KEY IS UniqueRecKey
ALT ERNAT ERECORD KEY IS AltKey WIT HDUP LICAT ES
FILE ST AT US IS FileStatus
 READ statement

READ FileName RECORD INTO DestItem

KEY IS KeyName
INVALID KEY StatementBlock
END READ

READ FileName NEXT RECORD INTO DestItem

AT END StatementBlock
END READ
 WRITE & REWRITE statement

WRITE RecName FROM SourceItem

INVALID KEY StatementBlock
END WRITE

REWRITE RecName FROM SourceItem

INVALID KEY StatementBlock
END REWRITE
 KSDS

 ORGANIZATION IS INDEXED
 ACCESS MODE IS SEQUENTIAL

INPUT OUTPUT INPUT-

OUTPUT
READ YES - YES

WRITE - YES

DELETE YES

REWRITE YES

START YES YES

INDEX FILE

ORGANIZATION IS INDEXED.
ACCESS MODE = SEQUENTIAL

OVERVIEW
CREATING KSDS
WRITING A KSDS
WRITING A KSDS
WRITING KSDS
RUNNING A KSDS
Running a KSDS
READING INDEX FILE-KSDS

ORGANIZATION IS INDEXED.
ACCESS MODE = SEQUENTIAL

OVERVIEW
READING A KSDS
READING A KSDS FILE
EXECUTE A PROGRAM FOR READING DATA
FROM KSDS
SPOOL OUTPUT
SPOOL OUTPUT
 START & DELETE statement

IS EQUAL TO
IS =
IS GREATER THAN
START FileName KEY KeyName
IS >
IS NOT LESS THAN
IS NOT <
INVALID KEY StatementBlock
END START

DELETE FileName RECORD

INVALID KEY StatementBlock
END DELETE
START :Position a Record
START COMMAND
START COMMAND
DELETE COMMAND
DELETE COMMAND
EXECUTE A PROGRAM AFTER COMPILE
PRINT THE KSDS DATASET
PRINT THE KSDS
KSDS
ORGANIZATION IS INDEXED
ACCESS MODE IS RANDOM
 KSDS

 ORGANIZATION IS INDEXED
 ACCESS MODE IS RANDOM

INPUT OUTPUT INPUT-

OUTPUT
READ YES - YES

WRITE - YES YES

DELETE YES

REWRITE YES

START
KSDS – RANDOM- WRITE
KSDS – Random -WRITE
KSDS –RANDOM MODE
READ Operation

Overview
KSDS – RANDOM -READ
KSDS – RANDOM -READ
KSDS
ORGANIZATION IS INDEXED
ACCESS MODE IS DYNAMIC
 KSDS

 ORGANIZATION IS INDEXED
 ACCESS MODE IS DYNAMIC

INPUT OUTPUT INPUT-

OUTPUT
READ YES - YES

WRITE - YES YES

DELETE YES

REWRITE YES

START YES YES

KSDS- DYNAMIC MODE
WRITE OPERATION

Overview
KSDS –DYNAMIC- WRITE
KSDS –DYNAMIC-WRITE
KSDS –DYNAMIC-READ
KSDS-DYNAMIC-READ
RRDS

Overview
 Sample JCL to create an RRDS cluster

//LEM0UXXA JOB
MSGCLASS=Q,CLASS=Q,NOTIFY=&SYSUID

//DEFRRDS EXEC PGM=IDCAMS NUMBERED - RRDS

//SYSPRINT DD SYSOUT=*

//SYSIN DD *

DEFINE CLUSTER(NAME(LEM0UXX.TEST.RRDS) -
RECORDSIZE(N1,N2)
VOLUMES(LEM001) -
N1=> AVG RECL
NUMBERED -
N2=> MAX RECL
TRACKS(2,1) -

RECORDSIZE(50,50) -

CONTROLINTERVALSIZE(4096) -
Freespace(PCT1,PCT2)
FREESPACE(10,20)) -
PCT1=> PCT FREE SPACE IN each CI
DATA(NAME(LEM0UXX.TEST.RRDS.DATA))
PCT2=> PCT Of unused CI in CA
/*
RELATIVE-RANDOM-WRITE
RELATIVE-RANDOM-WRITE
RELATIVE-RANDOM-READ
RELATIVE-RANDOM-READ
RANDOM READ JCL
SPOOL OUTPUT
RELATIVE-DYNAMIC-READ
RELATIVE-DYNAMIC-READ
RELATIVE-DYNAMIC-READ
ALTERNATE INDEX

Overview
 ALTERNATE INDEXES

 Used whenever the data is required to be retrieved on the basis of m

than one field
EMPNO ENAME SALARY
101 RAJESH 5000
102 RAMESH 6000
103 RANDY 7000
104 SURESH 8000

 e.g., Can be defined for both KSDS & ESDS

EMPNO IS ENAME IS
BASE KEY ALTERNATE KEY
 ALTERNATE INDEXES

Reduce data redundancy

Can have duplicates
Easy to define using IDCAMS
Allow datasets to be accessed sequentially or
randomly
Can be updated automatically
 Disadvantages

 Performance degradation
 Complex update logic
 Steps for Creating Alternate Index

 Define AIX using IDCAMS DEFINE AIX

 Specify Alternate Index Path using IDCAMS DEFINE PATH
 Build AIX & populate it using IDCAMS BLDINDEX
 Sample JCL to create an AIX cluster
UPGRADE-> VSAM AUTOMATIC

Updates the AIX for all

ADD,UPDT,DEL of Base cluster

//LEM0UXXA JOB
MSGCLASS=Q,CLASS=Q,NOTIFY=&SYSUID
//DEFKSDS EXEC PGM=IDCAMS
UNIQUE /NONUNIQUE
//SYSPRINT DD SYSOUT=*
//SYSIN DD *
DEFINE AIX (NAME(LEM0UXX.TEST.AIX.CLUSTER) –
KEY ( N1,N2)
RELATE (LEM0UXX.INFILE.KSDS) - N1-> LENGTH OF AIX key
VOLUMES(LEM001) -
KEY (10, 35) - N2-> KEYS start Loc in BASE CLUSTER
TRACKS(2,1) -
NONUNIQUEKEY - RECORDSIZE(N1,N2)
RECORDSIZE(49,49) -
N1=> AVG RECL
UPGRADE
CONTROLINTERVALSIZE(4096) - N2=> MAX RECL
FREESPACE(10,20)) -
DATA(NAME(LEM0UXX.TEST.AIX.DATA))
INDEX(NAME(LEM0UXX.TEST.AIX.INDEX)) Freespace(PCT1,PCT2)
/*
PCT1=> PCT FREE SPACE IN each CI

PCT2=> PCT Of unused CI in CA

 UPGRADE/NOUPGRADE

• Syntax : UPGRADE/NOUPGRADE

• UPGRADE specifies that records

in AIX are to be updated
• automatically whenever the
base cluster is updated
BUILDING AN INDEX

Overview
 BUILD INDEX

//JOB CARD
//STEP1 EXEC PGM=IDCAMS
//SYSPRINT DD SYSOUT=*
BUILD INDEX is used
//SYSIN DD *
To load records in
BLDINDEX -
INDATASET(LEM0UXX.KSDS.CLUSTER') - AIX after it has been def
OUTDATASET('LEM0UXX.KSDS.AIX.CLUSTER') -
/*
PATH AND ALTERNATE
INDEX

Overview
 DEFINE PATH

//JOB CARD
//STEP1 EXEC PGM = IDCAMS Path Establishes
//SYSPRINT DD SYSOUT=* A Bridge Between

//SYSIN DD * BASE CLUSTER & AIX

DEFINE PATH -
( NAME ( LEM0UXX.KSDS.PATH) –
PATHENTRY(LEM0UXX.KSDS.AIX.CLUSTER) –
UPDATE )
/*
//
ALTERNATE Index
Example
Define BASE CLUSTER
LOADING DATA INTO BASE CLUSTER
LOADING DATA INTO BASE CLUSTER
LOADING DATA INTO BASE CLUSTER
EXECUTING A PROGRAM TO LOAD DATA
EXECUTING A PROGRAM TO LOAD DATA
DEFINING AIX

RECORDSIZE(N1,N2)

N1= LENGTH OF BASE CLUSTER KEY(3)

+ LENGTH OF AIX KEY(5) + 5 for CI

Build index
Create path for Bridge
READING RECORDS BY ALTERNATE INDEX
READING RECORD BY ALTERNATE KEY
READING RECORD BY ALTERNATE KEY
EXECUTE A READ PROGRAM BY ALTERNATE
KEY

DD1-> BASE CLUSTER

DD11-> PATH NAME

DDNAME SHOULD BE IN ORD

SPOOL OUTPUT
IDCAMS REPRO
IDCAMS REPRO KEY BASED
IDCAMS REPRO RELATIVE
IDCAMS LISTCAT
SPOOL OUTPUT
IDCAMS DELETE