IMS-DB

Information Management System

DB Concepts & Programming
 Basic Topics
• Files Vs Databases
• Networked,Hierarchical & Relational structure
• Access types - HSAM, HISAM,HDAM,HIDAM,VSAM
• IMS DB - an Hierarchical Structure
• Database Definitions - DBD
• Program Specification Block ,Program Control Block
• DL/I Calls - Data Retrieval and Manipulation
• DL/I Functions
• Qualified/ Unqualified Calls, SSAs, Command Codes
• DL/I Status codes
 Advanced Topics
• Data Sensitivity - Segment Level & Field level
• Checkpoint & Restart
• IMS DB in CICS Environment
• Secondary Indexing
• Difference between MPP, BMP, DLI
• Using ‘HFCTDLI’
 File Vs Databases
• Data more redundant Redundancy avoided
• Duplicate data Minimise Duplicates
• Data insecure Security provided
• No control over data Field level sensitivity
• Threat to Integrity Consistency of data
• Lack of inbuilt Data Logging,
recovery features Commits & Rollbacks
• Lack of data Locks Automatic locks
 Database Structures
Network Structures (e.g IDMS)
• Many to Many relationships
• Complex structure using Pointers
• Best for Keeping storage
Hierarchical Structures (e.g IMS DB)
• One to Many Relationships
• Inverted Tree like Structure
• Best for storing parent and child type entities
Relational Structures (e.g DB2)
• One to Many, Many to One
• Best for ease of data retrieval
 Access Types
• HSAM/SHSAM (Simple /Hierarchic sequential Access Method)
The segments that make up the database record are related to one another by
their physical locations (SHSAM has only root occurrences)
• HISAM/SHISAM (Simple / Hierarchic Indexed Sequential Access Method)
The data is stored with hierarchic sequential organization along with an index
for random processing (SHISAM has only index of root occurrences)

• HDAM (Hierarchic Direct Access Method)

The segment occurrences include prefixes that contain the direct address
pointers to related segment - stores root segment occurrences
• HIDAM (Hierarchic Indexed Direct Access Method)
Similar to HDAM but stores an index of the root segment occurrences
 IMS DB Hierarchical Structure
Kanbay
Information KBunits Stratergic Business Units info
Database

Client Info Clients Employ Employee data

Project Info Projects Salary Employee Salary Details

 Basic IMS Terminologies
3
2
Kanbay 1
Information KBUnits Segment
Database Occurrence
Root Segment
Parent Segment
Child Segment
Client Info Clients Employ Segment Level
Path
Database Record
Twins
Project Info Projects Salary Sequential Processing
Random Processing
 Definitions
• Segment
- Unit of Information handled in IMS
- Unit of data transferred from DL/I in an IO operation
- Consists of fields, key fields & Search fields
• Occurrence
- Specific segment of a particular type of segment
- Contains actual user data
• Root Segment
- Top of the hierarchy
- All other segments are derived from the root
• Parent Segment
- Segment that has dependent segments
 Definitions (contd)
• Child Segment
- A dependent segment
• Segment Level
- Depth of an IMS structure
- Root is at Level 0, max 15 levels, max 255 segments
• Path
- Series of segments from Root to the child
• Database record
- Each occurrence of the Segment Plus all
occurrence of the subordinate segments
• Twins
- Segment occurrences of same type with same parent
segment occurrence
 IMS Control Blocks
PCB for DB1 IMS IMS
PSB PGM1 PGM2
PCB for DB2
PCB for DB3

DBDs DBD1 DBD2 DBD3

DL/I
CONTROL
UNIT

IMS DBs
DB1 DB2 DB3
 DBD : Database Description, describes physical Database
DBDGEN : DBDGEN Utility, Describes the physical structure of
the Database and creates DBD

• DBD : Names the database and identifies the access method

DBD Name = KANBAY, ACCESS=(HIDAM,VSAM)
• DATASET : Specifies DD Name used in the JCL and other parameters
DATASET DD1=DDSAT,DEVICE=3380,SIZE=(4096),FRSPC=(0,15)
• SEGM : Names the segment. Identifies the parent and establishes the length
SEGM Name =KBU,PARENT=0,BYTES=40
• LCHILD : Names the logical segment used in secondary Indexing and Logical
database
LCHILD NAME=(KBUNAME,KANBAY1),PTR=INDX
• FIELD : Defines fields. And identifies search &key fields by SEQ parameter
FIELD NAME=(KBUCODE,SEQ,U)BYTES=3,START=1,TYPE=C
• DBDGEN: Indicates the end of the DBDGEN process
DBDGEN
• COMMENT: ‘*’ IN 3rd COLUMN for writing user comments
 Sample DBD Listing for Kanbay Info Database

DBD NAME=KANBAY,ACCESS=(HIDAM,VSAM)
DSG1 DATASET DD1=DDSAT,DEVICE=3380,SIZE=4096,FRSPC=0,15)
SEGM NAME=KBU,PARENT=0,BYTES=40
FIELD NAME=(KBUCODE,SEQ,U),BYTES=3,START=1,TYPE=P
FIELD NAME=KBUNAME,BYTES=5,START=4,TYPE=C
FIELD NAME=KBULOC,BYTES=12,START=9,TYPE=C
FIELD NAME=KBUHEAD,BYTES=20,START=21,TYPE=C
SEGM NAME=CLIENT,PARENT=KBU,BYTES=45
FIELD NAME=(CLTID,SEQ,U),BYTES=4,START=1,TYPE=P
FIELD NAME=CLTNAME,BYTES=10,START=5,TYPE=C
FIELD NAME=CLTLOC,BYTES=11,START=15,TYPE=C
FIELD NAME=CLTHEAD,BYTES=20,START=26,TYPE=C
SEGM NAME=PROJECT,PARENT=CLIENT,BYTES=35
FIELD NAME=(PRJID,SEQ,U),BYTES=4,START=1,TYPE=P
FIELD NAME=PRJNAME,BYTES=5,START=5,TYPE=C
FIELD NAME=PRJTMSIZE,BYTES=4,START=10,TYPE=C
FIELD NAME=PRJHEAD,BYTES=22,START=14,TYPE=C
SEGM NAME=EMPLY,PARENT=KBU,BYTES=40
FIELD NAME=(EMPID,SEQ,U),BYTES=5,START=1,TYPE=P
FIELD NAME=EMPNAME,BYTES=10,START=6,TYPE=C
FIELD NAME=EMPPROJ,BYTES=5,START=16,TYPE=C
FIELD NAME=EMPADDR,BYTES=20,START=21,TYPE=C
SEGM NAME=SALARY,PARENT=EMPLY,BYTES=35
FIELD NAME=(SALEMPID,SEQ,U),BYTES=5,START=1,TYPE=P
FIELD NAME=SALBASIC,BYTES=10,START=6,TYPE=C
FIELD NAME=SALHRA,BYTES=5,START=16,TYPE=C
FIELD NAME=SALDA,BYTES=5,START=21,TYPE=C
DBDGEN
FINISH
END
PSB
Program Specification Block,describes Logical structure of the Database that a
program will use.It contains one or more PCB’s (Program Communication Block).
PCB describes one Database.

PSBGEN
PSBGEN Produces a load module stored in the PSBLIB library and creates PSB
Major Macros Used
PCB : Describes type (DB or DC), name of the DBD, Key length(length of
the longest concatenated key the program can process) and
processing Option (Defines how the segment may be accessed)
PCB TYPE=DB,DBDNAME=KANBAY,KEYLEN=20,PROCOPT=LS
SENSEG : Define the name of the segments in the Database your program
needs access to
SENSEG NAME=CLIENT,PARENT=KBU
SENFLD : Subordinate to the SENSEG macro, only fields that are defined by
the SENFLD macro can be accessed from the I/O Area.
SENFLD NAME=PROJECT,PARENT=CLIENT
PSBGEN PSBNAME=KANPSB,LANG=COBOL
END
SAMPLE PSB FOR KANBAY INFO DB PROJECT REFERENCE PROGRAM

PCB TYPE=DB,DBDNAME=KANBAY,KEYLEN=10,PROCOPT=LS
SENSEG NAME=KBU,PARENT=0
SENSEG NAME=CLIENT,PARENT=KBU
SENSEG NAME=PROJECT,PARENT=CLIENT
PSBGEN PSBNAME=KANPSB,LANG=COBOL
END

SAMPLE PSB FOR KANBAY INFO DB PROJECT REFERENCE PROGRAM

PCB TYPE=DB,DBDNAME=KANBAY,KEYLEN=10,PROCOPT=LS
SENSEG NAME=KBU,PARENT=0
SENSEG NAME=CLIENT,PARENT=KBU
SENSEG NAME=PROJECT,PARENT=CLIENT
SENSEG NAME=EMPLY,PARENT=KBU
SENSEG NAME=SALARY,PARENT=EMPLY
PSBGEN PSBNAME=KANPSB,LANG=COBOL
END
 DL/I INTERFACE

• DL/I passes control to your program

• Uses ENRTY and CALL statement to code DL/I call in the Procedure Division

ENTRY ‘DLITCBL’ USING PCB-name1

PCB-name2

This DL/I tells entry of the program for Batch program & MPP program

CALL ‘CBLTDLI’ USING DLI-FUNCTION

PCB-mask
Segment-io-area
[segment search arguments (s)]

This call is the name of the interface module that link edited with your
program’s object module.
 CALL PARAMETERS

CALL ‘CBLDTLI’ USING DLI-FUNCTION

PCB-mask
Segment-io-area
[segment search argument (s) ]

If we use PL/I language then CBLTDLI will be PLITDLI

Parameter Function

DL/I -Function A four byte code that tells DL/I what kind of call to make
PCB-mask Arguments tells DL/I which PCB to use for the call
IO-Area Tells DL/I where to put the segment
SSA Gives DL/I additional criteria to use in finding a particular
Segment, Can be upto 15 SSA’s
 DL/I FUNCTIONS

GU - Get Unique
GN - Get Next
GNP - Get Next within Parent
GHU - Get Hold Unique
GHN - Get Hold Next
GHNP - Get Hold Next within Parent
ISRT - To Insert
REPL - To replace
DLET - To Delete
CHKP - To take a Checkpoint at a particular point
XRST - To Restart from Particular point previous Checkpoint
INIT - To Initialize
SYNC - To commit
 PCB MASK

Define in linkage Section and establishes correspondence between linkage Section

PSB-mask and PSB’s with in your program.

SAMPLE PCB-MASK LAYOUT

VARIABLE DESCRIPTION

01 KANBAY-PCB-MASK.
03 KAN-DBD-NAME PIC X(8). Name of the Database being processed
03 KAN-SEG-LEVEL PIC XX. Specifies current segment level in the DB
03 KAN-STATUS-CODE PIC XX. Contains data yr program will evaluate after call
03 KAN-PROC-OPT PIC X(4). Type of processing the program authorized to do
03 KAN-SEG-NAME PIC X(8). Stores name of the segment just proceeded
03 KAN-KEY-LENGTH PIC S9(5). Length of the concatenated key
03 KAN-NUM-SENSEGS PIC S9(5). No of SENSEG macros subordinate to the PCB
03 KAN-KEY-VALUE PIC X(20). Key value retrieved
 I-O AREA LAYOUT FOR SEGMENTS

KBUCODE KBUNAME KBULOC KBUHEAD

01 KBU-AREA.
03 KBU-CODE PIC X(3). EMPID EMPNAME EMPPROJ EMPADR
03 KBU-NAME PIC X(5).
03 KBU-LOC PIC X(12) 01 EMP-AREA.
03 KBU-HEAD PIC X(20). 03 EMP-ID PIC X(5).
03 EMP-NAME PIC X(10).
03 EMP-PROJ PIC X(5)
CLTID CLTNAME CLTLOC CLT-HEAD
03 EMP-ADR PIC X(20).
01 CLIENT-AREA.
03 CLT-CLTID PIC X(4).
EMPID BASIC HRA DA
03 CLT-CLTNAME PIC X(10).
03 CLT-CLTLOC PIC X(11)
03 CLT-CLTHEAD PIC X(20). 01 SAL-AREA.
03 SAL-EMPID PIC X(5).
PROJID PRJNAME TMSIZE PRJADDR 03 SAL-BASIC PIC X(10).
01 PRJ-AREA. 03 SAL-HRA PIC X(5).
03 PRJ-PROJID PIC X(4). 03 SAL-DA PIC X(5).
03 PRJ-PRJNAME PIC X(5).
03 PRJ-TMSIZE PIC X(4)
03 PRJ-HEAD PIC X(20).
 SEGMENT SEARCH ARGUMENT
• Identifies Segment Occurrence You want to access
• You may have to code several SSA’s in a single DL/I CALL
• There are two types of SSA’s Unqualified SSA & Qualified SSA

UNQUALIFIED QUALIFIED
• Search for a Specific Segment type Searches for a specific occurrence of a
specific segment type
• Access the database Sequentially Access the database Randomly
• Identify by segment name and space Identify by segment name and Seg-key

• Example
01 UNKBU-SSA. 01 QUALKBU-SSA.
03 KBU-UNSSA-NM PIC X(8) Value ‘KBU ‘. 03 SEG-NAME PIC X(8) VALUE ‘KBU ‘.
03 FILLER PIC X Value Space. 03 COM-CODE PIC X(2) VALUE ‘*-’.
03 BEGIN-QUAL PIC X VALUE ‘(‘.
03 KEY-NAME PIC X(8) VALUE ‘KBUKEY’.
03 REL-OPR PIC X(2) VALUE ‘= ‘.
03 KEY-VALUE PIC X(20).
03 END-QUAL PIC X VALUE ‘)’.
 COMMAND CODES

Command Code Description

D Retrieve Path
F Access First Segment occurrence
L Access Last Segment occurrence
U Maintain current position at this level
V Maintain current position at this level and higher levels
Q Enqueue Segment
N Do Not replace this Segment
C Use a Concatenated Key
P Establish Percentage at this level
- Null command code / ignore
 Data Sensitivity

•DBA
At a PSB level for SEGMENTS PROCOPT
PCB TYPE=DB,DBDNAME-KANBAY,,KEYLEN=11,PROCOPT
SENSEG NAME =KBU,PARENT=0
SENSEG NAME =CLIENT,PARENT=KBU

Processing Options : G, I, R, D, A, P, O, N, T, E, L, GS, LS, H

At a PSB level for fields

Done by DBA by use of SENFLD macro , specifying those fields
that can be changed.

•Programmer
At PROGRAM Level
- Qualified SSA call with Command code N (No Replacement)
 Checkpointing & Restart
• A point in the program where the database changes the program made are
Considered complete and accurate -written to a DL/I log
• Used for Recovery Process
• Are of two types - Basic Checkpointing and Symbolic Checkpointing
• Basic Checkpointing
CALL ‘CBLTDLI’ USING DLI-CHKP,
I-O-PCB-MASK,
CHECKPOINT-ID.

• Symbolic Checkpointing
Stores program data along with checkpoint records and retrieves that data
when program restarts after failure.

For extended Restart , an XRST call to be issued with the same fields as in
CHKP call. It allows DL/I to retrieve the value in checkpoint record and
restores the specified fields.
 Explosive Mix - IMS DB with CICS
• Scheduling Call

CICS Region / Partition

User Interface Block

(UIB)

User Task

Program Communication
Block
(PCB)

A DL/I task must establish addressability to UIB and one or more PCB
 IMS & CICS Continued

CALL ‘CBLTDLI’ USING FUNC-INIT User Interface Block

PSB-NAME
ADDRESS OF DLIUIB
Address of PCB List

SET ADDRESS OF PCB-ADDR-LIST TO UIBPCBAL

SET ADDRESS OF PCB-1 TO PCB-1-ADR Address of PCB 1
SET ADDRESS OF PCB-2 TO PCB-2-ADR
…... PCB Address list
Address of PCB 2

PCB1

PCB2
 Secondary Indexing

Employee Index Database

EMP..Seg Employee ID
Addr
Index Target
KBU Index Pointer Segment
Segment

Index Source Segment

CLIENTS EMPLOYEE

Same index Source segment and index target segment

PROJECTS
 Secondary Indexing Continued

Employee Index Database

KBU..Seg Employee ID
Addr
Index Target
KBU Index Pointer Segment
Segment

Index Source Segment

CLIENTS EMPLOYEE

Different index Source segment and index target segment

PROJECTS
 Secondary Indexing Continued

• Alternate sequencing is provided by storing pointers to Segments of the indexed

Database in a separate secondary index data base

• Secondary index data base has only one segment type - the Index pointer segment
which in turn contains a prefix element and a data element

• Prefix element points to the index target segment

• Secondary indexing is useful for accessing segments were both

Index source segment and index target segment are same

Index source segment and index target segment are different

• Two DBDGENS are required as a secondary database involves two database: one
for indexed database , the other for the secondary index database
 Secondary Indexing Continued

DBDGEN - The Secondary Index code for Kanbay DB

DBD NAME=KANBAY,ACCESS=(HIDAM,VSAM)
DSG1 DATASET DD1=DDSAT,DEVICE=3380,SIZE=4096,FRSPC=0,15)
SEGM NAME=KBU,PARENT=0,BYTES=40
FIELD NAME=(KBUCD,SEQ,U),BYTES=2,START=1,TYPE=P
……
SEGM NAME=EMPLY,PARENT=KBU,BYTES=80
LCHILD NAME= (KANPNTR,KANDBD),POINTER=INDX
FIELD NAME=(EMPID,SEQ,U),BYTES=5,START=1,TYPE=P
XDFLD NAME=KANINDX,SRCH=EMPID
FIELD NAME=EMPNAME,BYTES=25,START=6,TYPE=C
FIELD NAME=EMPPROJ,BYTES=15,START=31,TYPE=C
FIELD NAME=EMPADDR,BYTES=35,START=46,TYPE=C
…….
 Secondary Indexing Continued

DBDGEN - The Secondary Index code for Secondary Index DB

DBD NAME=KANIDBD,ACCESS=(INDEX)
DSG1 DATASET DD1=DDKAY,DEVICE=3380,SIZE=4096,FRSPC=0,15)
SEGM NAME=KANPNTR,PARENT=0,BYTES=6
LCHILD NAME= (EMPID,KANBAY), INDX=KANINDX
FIELD NAME=(KANEMPID,SEQ,U),BYTES=6,START=1,TYPE=C
DBDGEN
FINISH
END

PSB Requirements for Secondary Indexing.

PCB TYPE=DB,DBDNAME=KANBAY,KEYLEN=10,PROCOPT=LS,
PROCSEQ=KANIDBD
SENSEG NAME=EMPLY
SENSEG NAME=KBU,PARENT=EMPLY
PSBGEN PSBNAME=KANSBX,LANG=COBOL
END
 Difference Between MPP , BMP & DLI

• MPP (Message Processing Program)

For Online programs( IMSDC) only.
PSB used by a MPP program cannot be used by another programs simultaneously.
Cannot run on Individual partitions.

• BMP (Batch message processing)

For batch programs only
PSB used by a BMP program cannot be used by another programs simultaneously.
All partitions specified in PSB will be allocated whether used by program or not.
Cannot run on Individual partitions.

• DL/I
It can be used as Interface between MPP and Database OR BMP and Database
For batch DL/I job it can access database directly
PSB of DL/I job can be used by another jobs simultaneously
It can run on Individual partitions.
 Using HFCTDLI

HFCTDLI: Is an House Hold utility which is used instead of CBLTDLI

As House Hold is having same data of particular segment in separate

partitions (I.e on different files)

So while accessing data from segment (I.e DBD) through program

HFCTDLI is used ,to treat all the partitions as one.