Ibm Mainframe

IBM Mainframe

Overview

Comprehensive guidance for IBM Mainframe (z/OS) administration, development, operations, and modernization. Covers system programming, application development, middleware configuration, database management, and migration strategies for mainframe environments.

Quick Start Guide

Choose your task and load the appropriate reference:

1. z/OS System Administration → Load zos-administration.md for system operations, address spaces, JES2/JES3, TSO/ISPF commands, dataset management, and USS 2. JCL Development & Batch Processing → Load jcl-batch-processing.md for JCL syntax, utilities (IEBGENER, SORT, IDCAMS), procedures, GDG management, and job scheduling 3. COBOL Programming → Load cobol-programming.md for COBOL structure, file handling, DB2/CICS integration, compile/link procedures, and debugging 4. Mainframe Modernization → Load mainframe-modernization.md for migration strategies, assessment, rehost/replatform/refactor approaches, and cloud integration

Core Mainframe Concepts

z/OS Operating System

• MVS (Multiple Virtual Storage): Core OS components
• JES2/JES3: Job Entry Subsystems for batch processing
• TSO/ISPF: Time Sharing Option / Interactive System Productivity Facility
• USS: Unix System Services (POSIX environment on z/OS)
• Address Spaces: Isolated execution environments (ASID)

Key Subsystems

• CICS: Customer Information Control System (transaction processing)
• IMS: Information Management System (hierarchical DB + transaction manager)
• DB2: Relational database management system
• MQ: Message queuing middleware
• WebSphere: Application server for Java applications

Data Management

• VSAM: Virtual Storage Access Method (indexed, sequential, relative record)
• Sequential Files: PS (Physical Sequential), PDS (Partitioned Data Set), PDSE
• GDG: Generation Data Group (versioned datasets)
• SMS: Storage Management Subsystem (automated data management)

Security Systems

• RACF: Resource Access Control Facility (IBM)
• ACF2: Access Control Facility 2 (CA Technologies)
• Top Secret: Security system (CA Technologies)

Common Mainframe Tasks

For detailed syntax, examples, and procedures, load the appropriate reference file:

Administration & Operations

• JCL job submission & monitoring: Load jcl-batch-processing.md
• Dataset management: Load zos-administration.md
• VSAM file operations: Load jcl-batch-processing.md

Application Development

• COBOL programming: Load cobol-programming.md
• CICS transaction management: Load cobol-programming.md
• DB2 operations: Load cobol-programming.md

System Management

• Security (RACF/ACF2): Load zos-administration.md
• Performance monitoring: Load zos-administration.md

Mainframe Development Workflow

For complete development procedures, compile JCL, version control integration, and testing strategies, load cobol-programming.md.

Typical development cycle:

1. Edit source code (ISPF or modern IDE) 2. Compile and link programs 3. Test in development environment 4. Version control and promotion 5. Deploy to production

Modern development approaches:

• Git integration via USS (Unix System Services)
• Zowe CLI for remote access and CI/CD
• DBB (Dependency Based Build) for automated builds
• Eclipse/VS Code with mainframe extensions

Mainframe Modernization Strategies

For comprehensive migration guidance, assessment procedures, and specific migration patterns, load mainframe-modernization.md.

Key approaches:

1. Rehost (Lift & Shift): Mainframe emulators on x86/cloud (fastest, minimal changes) 2. Replatform: Migrate to modern languages (Java, C#) with automated conversion 3. Refactor: Redesign as microservices with modern architecture patterns 4. Retire/Replace: Replace with COTS or modern SaaS alternatives

Integration options:

• z/OS Connect for RESTful APIs
• IBM MQ for messaging
• CDC (Change Data Capture) for data replication
• DevOps integration with Zowe, Jenkins, Git

Best Practices

Development

1. Use structured programming (avoid ALTER, computed GO TO) 2. Modular design with focused paragraphs/sections 3. Comprehensive error handling (EXEC CICS HANDLE, SQL SQLCODE checks) 4. Clear documentation with meaningful variable names 5. Performance optimization (minimize I/O, efficient SQL, proper indexing)

Operations

1. Automation with JCL procedures, REXX scripts, automation tools 2. Proactive monitoring with alerts and performance baselines 3. Regular backup & recovery with DR testing 4. Capacity planning to monitor trends and forecast growth 5. Security with least privilege, regular audits, compliance

Migration

1. Phased approach to reduce risk incrementally 2. Data integrity validation before and after migration 3. Parallel running of old and new systems 4. Maintain rollback plans 5. Extensive documentation and team training

For detailed procedures, load the appropriate reference file.

Tooling Ecosystem

IBM Native Tools

• ISPF: Primary development interface
• SDSF: Job and output management
• RMF/SMF: Performance monitoring and system logging
• OMVS: Unix shell environment on z/OS

Modern Mainframe Tools

• Zowe: Open-source framework for mainframe modernization
• DBB: Dependency Based Build for DevOps
• IBM Wazi: Cloud-native development environment
• Topaz/IBM Developer for z/OS: Modern IDEs

Migration Tools

• Micro Focus Enterprise Suite: Rehost solution
• LzLabs/AWS Mainframe Modernization: Cloud migration platforms
• IBM Rational/CAST: Application analysis tools

For detailed tool usage, configuration, and examples, load the appropriate reference file.

Common Issues & Solutions

Job Failures

• S0C7 (Data exception): Invalid numeric data, uninitialized variables, data conversion errors
• S806 (Program not found): Missing STEPLIB/JOBLIB, linking issues, wrong load library
• S322 (Time out): Infinite loop, long-running query, insufficient TIME parameter
• S013/S213/S413 (OPEN failure): Dataset allocation issues, DISP conflicts, missing dataset

Performance Issues

• High CPU: Inefficient code, missing indexes, excessive I/O
• Long response time: Network latency, database contention, poor SQL
• Memory constraints: Virtual storage shortage, insufficient region size

Security Issues

• Access denied: Check RACF permissions and group memberships
• Unauthorized program: Not in APF library or not properly authorized
• Dataset protection: Review UACC and specific permits

For detailed troubleshooting procedures, load the appropriate reference file.

Key Concepts Glossary

• DASD: Direct Access Storage Device (disk storage)
• ABEND: Abnormal End (program crash)
• IPL: Initial Program Load (system boot)
• LPAR: Logical Partition (virtual machine)
• SYSPLEX: System Complex (cluster of z/OS systems)
• Coupling Facility: High-speed memory for parallel sysplex
• HCD/HCM: Hardware Configuration Dialog/Manager
• SMP/E: System Modification Program/Extended (maintenance)
• ISPF: Interactive System Productivity Facility
• REXX: Restructured Extended Executor (scripting language)

Critical Considerations

Data Integrity

• VSAM files require proper SHAREOPTIONS
• DB2 locking and isolation levels must be configured correctly
• GDG generation management for versioned datasets
• Always backup before changes

System Performance

• Batch window constraints (limited time for batch jobs)
• Online transaction response time requirements (sub-second)
• Resource contention management (CPU, I/O, memory)
• Capacity planning for growth

Security & Compliance

• RACF/ACF2 configuration and maintenance
• Audit trails and compliance requirements (SOX, PCI-DSS, GDPR)
• Separation of duties enforcement

High Availability

• Parallel Sysplex for redundancy
• GDPS (Geographically Dispersed Parallel Sysplex) for disaster recovery
• Continuous availability requirements

Output Requirements

When analyzing or documenting mainframe systems, provide:

1. System inventory: Operating system version, middleware, applications 2. Architecture diagram: LPAR layout, subsystems, connectivity 3. Dependency mapping: Job dependencies, program CALL trees, data flows 4. Security assessment: User access, dataset protection, compliance gaps 5. Performance baseline: CPU, I/O, response times, throughput 6. Modernization roadmap: Assessment results, approach, timeline, risks (load mainframe-modernization.md) 7. Migration plan: Phased approach, cutover strategy, rollback procedures 8. Documentation: Technical specifications, runbooks, operational procedures

For specific output formats and templates, load the appropriate reference file.

Next Steps

For specific tasks, load the relevant reference file from the Quick Start Guide above. Each reference provides detailed procedures, commands, examples, and troubleshooting guidance for that specific area of mainframe operations.

COBOL Programming on Mainframe

COBOL Program Structure

Complete Program Template

       IDENTIFICATION DIVISION.
       PROGRAM-ID. CUSTINQ.
       AUTHOR. DEVELOPMENT TEAM.
       DATE-WRITTEN. 2024-01-15.
      *****************************************************************
      * PROGRAM: CUSTINQ - Customer Inquiry                          *
      * PURPOSE: Query customer data from DB2                        *
      * INPUT:   Customer ID from terminal                           *
      * OUTPUT:  Customer details to terminal                        *
      *****************************************************************
       
       ENVIRONMENT DIVISION.
       CONFIGURATION SECTION.
       SOURCE-COMPUTER. IBM-Z15.
       OBJECT-COMPUTER. IBM-Z15.
       
       INPUT-OUTPUT SECTION.
       FILE-CONTROL.
           SELECT CUSTOMER-FILE
               ASSIGN TO CUSTMAST
               ORGANIZATION IS INDEXED
               ACCESS MODE IS RANDOM
               RECORD KEY IS CUST-ID
               FILE STATUS IS WS-FILE-STATUS.
       
       DATA DIVISION.
       FILE SECTION.
       FD  CUSTOMER-FILE
           RECORDING MODE IS F
           BLOCK CONTAINS 0 RECORDS.
       01  CUSTOMER-RECORD.
           05  CUST-ID                 PIC 9(10).
           05  CUST-NAME               PIC X(50).
           05  CUST-ADDRESS            PIC X(100).
           05  CUST-BALANCE            PIC S9(11)V99 COMP-3.
           05  CUST-STATUS             PIC X(01).
               88  ACTIVE-CUSTOMER     VALUE 'A'.
               88  INACTIVE-CUSTOMER   VALUE 'I'.
       
       WORKING-STORAGE SECTION.
       01  WS-FILE-STATUS              PIC XX.
           88  FILE-SUCCESS            VALUE '00'.
           88  EOF-REACHED             VALUE '10'.
           88  RECORD-NOT-FOUND        VALUE '23'.
       
       01  WS-COUNTERS.
           05  WS-RECORDS-READ         PIC 9(07) COMP VALUE ZERO.
           05  WS-RECORDS-WRITTEN      PIC 9(07) COMP VALUE ZERO.
       
       01  WS-DISPLAYS.
           05  WS-EDIT-BALANCE         PIC ZZZ,ZZZ,ZZ9.99-.
       
       PROCEDURE DIVISION.
       0000-MAIN-ROUTINE.
           PERFORM 1000-INITIALIZE
           PERFORM 2000-PROCESS-RECORDS
           PERFORM 3000-FINALIZE
           STOP RUN.
       
       1000-INITIALIZE.
           OPEN INPUT CUSTOMER-FILE
           IF NOT FILE-SUCCESS
               DISPLAY 'ERROR OPENING CUSTOMER FILE: ' WS-FILE-STATUS
               MOVE 16 TO RETURN-CODE
               STOP RUN
           END-IF
           DISPLAY 'INITIALIZATION COMPLETE'.
       
       2000-PROCESS-RECORDS.
           PERFORM 2100-READ-CUSTOMER
           PERFORM UNTIL EOF-REACHED
               PERFORM 2200-PROCESS-CUSTOMER
               PERFORM 2100-READ-CUSTOMER
           END-PERFORM.
       
       2100-READ-CUSTOMER.
           READ CUSTOMER-FILE NEXT
               AT END
                   SET EOF-REACHED TO TRUE
               NOT AT END
                   ADD 1 TO WS-RECORDS-READ
           END-READ.
       
       2200-PROCESS-CUSTOMER.
           IF ACTIVE-CUSTOMER
               MOVE CUST-BALANCE TO WS-EDIT-BALANCE
               DISPLAY 'CUSTOMER: ' CUST-NAME
               DISPLAY 'BALANCE:  ' WS-EDIT-BALANCE
               ADD 1 TO WS-RECORDS-WRITTEN
           END-IF.
       
       3000-FINALIZE.
           CLOSE CUSTOMER-FILE
           DISPLAY 'RECORDS READ:    ' WS-RECORDS-READ
           DISPLAY 'RECORDS WRITTEN: ' WS-RECORDS-WRITTEN
           DISPLAY 'PROCESSING COMPLETE'.

Data Types & Definitions

Numeric Types

      * Display numeric (EBCDIC)
       01  WS-AMOUNT-1         PIC 9(07)V99 VALUE ZERO.
       
      * Signed display numeric
       01  WS-AMOUNT-2         PIC S9(07)V99 VALUE ZERO.
       
      * Binary (COMP)
       01  WS-COUNTER          PIC 9(07) COMP VALUE ZERO.
       01  WS-INDEX            PIC S9(04) COMP VALUE ZERO.
       
      * Packed decimal (COMP-3) - Recommended for calculations
       01  WS-BALANCE          PIC S9(11)V99 COMP-3 VALUE ZERO.
       01  WS-TOTAL            PIC S9(13)V99 COMP-3 VALUE ZERO.

Character Types

      * Alphanumeric
       01  WS-NAME             PIC X(50) VALUE SPACES.
       01  WS-ADDRESS          PIC X(100) VALUE SPACES.
       
      * Alphabetic
       01  WS-ALPHA-FIELD      PIC A(30) VALUE SPACES.
       
      * National (Unicode)
       01  WS-UNICODE-NAME     PIC N(50) VALUE SPACES.

Group Items & Structures

       01  CUSTOMER-RECORD.
           05  CUST-KEY.
               10  CUST-TYPE           PIC X(02).
               10  CUST-ID             PIC 9(10).
           05  CUST-DATA.
               10  CUST-NAME           PIC X(50).
               10  CUST-ADDRESS.
                   15  ADDR-LINE-1     PIC X(40).
                   15  ADDR-LINE-2     PIC X(40).
                   15  ADDR-CITY       PIC X(30).
                   15  ADDR-STATE      PIC X(02).
                   15  ADDR-ZIP        PIC 9(05).
           05  CUST-FINANCIAL.
               10  CUST-BALANCE        PIC S9(11)V99 COMP-3.
               10  CUST-CREDIT-LIMIT   PIC S9(09)V99 COMP-3.

Arrays (OCCURS)

      * Fixed array
       01  MONTHLY-SALES.
           05  MONTH-SALE          PIC S9(09)V99 COMP-3
                                   OCCURS 12 TIMES.
       
      * Variable array
       01  TRANSACTION-TABLE.
           05  TXN-COUNT           PIC 9(04) COMP.
           05  TXN-ENTRY           OCCURS 1 TO 9999 TIMES
                                   DEPENDING ON TXN-COUNT.
               10  TXN-DATE        PIC 9(08).
               10  TXN-AMOUNT      PIC S9(09)V99 COMP-3.
       
      * Indexed array
       01  PRODUCT-TABLE.
           05  PRODUCT-ENTRY       OCCURS 500 TIMES
                                   INDEXED BY PROD-IDX.
               10  PROD-CODE       PIC X(10).
               10  PROD-DESC       PIC X(50).
               10  PROD-PRICE      PIC S9(07)V99 COMP-3.

Condition Names (88 Levels)

       01  RECORD-STATUS           PIC X(01).
           88  STATUS-ACTIVE       VALUE 'A'.
           88  STATUS-INACTIVE     VALUE 'I'.
           88  STATUS-PENDING      VALUE 'P'.
           88  STATUS-CLOSED       VALUE 'C'.
           88  VALID-STATUS        VALUES 'A' 'I' 'P' 'C'.
       
       01  TRANSACTION-TYPE        PIC XX.
           88  DEBIT-TXN           VALUE '01' '02' '05'.
           88  CREDIT-TXN          VALUE '10' '11' '15'.

REDEFINES

       01  DATE-FIELD.
           05  DATE-NUMERIC        PIC 9(08).
           05  DATE-PARTS REDEFINES DATE-NUMERIC.
               10  DATE-YEAR       PIC 9(04).
               10  DATE-MONTH      PIC 9(02).
               10  DATE-DAY        PIC 9(02).
       
       01  ACCOUNT-KEY.
           05  ACCOUNT-NUMBER      PIC X(10).
           05  ACCOUNT-BREAKDOWN REDEFINES ACCOUNT-NUMBER.
               10  ACCT-BRANCH     PIC X(03).
               10  ACCT-TYPE       PIC X(02).
               10  ACCT-SEQUENCE   PIC X(05).

File Processing

Sequential File

       FILE-CONTROL.
           SELECT INPUT-FILE
               ASSIGN TO INFILE
               ORGANIZATION IS SEQUENTIAL
               ACCESS MODE IS SEQUENTIAL
               FILE STATUS IS WS-INPUT-STATUS.
       
       PROCEDURE DIVISION.
           OPEN INPUT INPUT-FILE
           PERFORM READ-RECORD
           PERFORM UNTIL EOF-INPUT
               PERFORM PROCESS-RECORD
               PERFORM READ-RECORD
           END-PERFORM
           CLOSE INPUT-FILE.
       
       READ-RECORD.
           READ INPUT-FILE
               AT END
                   SET EOF-INPUT TO TRUE
           END-READ.

Indexed File (VSAM KSDS)

       FILE-CONTROL.
           SELECT MASTER-FILE
               ASSIGN TO MASTER
               ORGANIZATION IS INDEXED
               ACCESS MODE IS RANDOM
               RECORD KEY IS MASTER-KEY
               ALTERNATE RECORD KEY IS MASTER-ALT-KEY
                   WITH DUPLICATES
               FILE STATUS IS WS-MASTER-STATUS.
       
       PROCEDURE DIVISION.
      * Random read
           MOVE '1234567890' TO MASTER-KEY
           READ MASTER-FILE
               INVALID KEY
                   DISPLAY 'RECORD NOT FOUND'
           END-READ.
       
      * Sequential read
           OPEN INPUT MASTER-FILE
           PERFORM READ-NEXT
           PERFORM UNTIL EOF-MASTER
               PERFORM PROCESS-RECORD
               PERFORM READ-NEXT
           END-PERFORM.
       
       READ-NEXT.
           READ MASTER-FILE NEXT
               AT END
                   SET EOF-MASTER TO TRUE
           END-READ.
       
      * Update
           MOVE '1234567890' TO MASTER-KEY
           READ MASTER-FILE UPDATE
           IF FILE-SUCCESS
               MOVE 'NEW VALUE' TO MASTER-DATA
               REWRITE MASTER-RECORD
           END-IF.
       
      * Write new record
           MOVE '9999999999' TO MASTER-KEY
           MOVE 'NEW CUSTOMER' TO MASTER-NAME
           WRITE MASTER-RECORD
               INVALID KEY
                   DISPLAY 'DUPLICATE KEY'
           END-WRITE.
       
      * Delete
           MOVE '1234567890' TO MASTER-KEY
           DELETE MASTER-FILE
               INVALID KEY
                   DISPLAY 'RECORD NOT FOUND'
           END-DELETE.

DB2 Integration

SQL Declaration

       DATA DIVISION.
       WORKING-STORAGE SECTION.
           EXEC SQL
               INCLUDE SQLCA
           END-EXEC.
           
           EXEC SQL
               INCLUDE CUSTOMER
           END-EXEC.
           
           EXEC SQL DECLARE CUSTCUR CURSOR FOR
               SELECT CUST_ID, CUST_NAME, CUST_BALANCE
               FROM CUSTOMER
               WHERE CUST_STATUS = 'A'
               ORDER BY CUST_NAME
           END-EXEC.
       
       01  WS-CUST-ID              PIC 9(10).
       01  WS-CUST-NAME            PIC X(50).
       01  WS-CUST-BALANCE         PIC S9(11)V99 COMP-3.

SQL Operations

       PROCEDURE DIVISION.
      * Single row SELECT
           EXEC SQL
               SELECT CUST_NAME, CUST_BALANCE
               INTO :WS-CUST-NAME, :WS-CUST-BALANCE
               FROM CUSTOMER
               WHERE CUST_ID = :WS-CUST-ID
           END-EXEC
           
           IF SQLCODE = 0
               DISPLAY 'CUSTOMER: ' WS-CUST-NAME
           ELSE
               IF SQLCODE = 100
                   DISPLAY 'CUSTOMER NOT FOUND'
               ELSE
                   DISPLAY 'SQL ERROR: ' SQLCODE
               END-IF
           END-IF.
       
      * Cursor processing
           EXEC SQL
               OPEN CUSTCUR
           END-EXEC
           
           PERFORM UNTIL SQLCODE NOT = 0
               EXEC SQL
                   FETCH CUSTCUR
                   INTO :WS-CUST-ID, :WS-CUST-NAME, :WS-CUST-BALANCE
               END-EXEC
               
               IF SQLCODE = 0
                   PERFORM PROCESS-CUSTOMER
               END-IF
           END-PERFORM
           
           EXEC SQL
               CLOSE CUSTCUR
           END-EXEC.
       
      * INSERT
           EXEC SQL
               INSERT INTO CUSTOMER
                   (CUST_ID, CUST_NAME, CUST_BALANCE)
               VALUES
                   (:WS-CUST-ID, :WS-CUST-NAME, :WS-CUST-BALANCE)
           END-EXEC.
       
      * UPDATE
           EXEC SQL
               UPDATE CUSTOMER
               SET CUST_BALANCE = :WS-CUST-BALANCE
               WHERE CUST_ID = :WS-CUST-ID
           END-EXEC.
       
      * DELETE
           EXEC SQL
               DELETE FROM CUSTOMER
               WHERE CUST_ID = :WS-CUST-ID
           END-EXEC.
       
      * COMMIT/ROLLBACK
           IF PROCESSING-OK
               EXEC SQL
                   COMMIT
               END-EXEC
           ELSE
               EXEC SQL
                   ROLLBACK
               END-EXEC
           END-IF.

CICS Programming

CICS Commands

       WORKING-STORAGE SECTION.
       01  WS-COMMAREA.
           05  CA-FUNCTION         PIC X(08).
           05  CA-CUST-ID          PIC 9(10).
           05  CA-CUST-DATA        PIC X(200).
       
       01  WS-RESP                 PIC S9(08) COMP.
       01  WS-RESP2                PIC S9(08) COMP.
       
       PROCEDURE DIVISION.
      * Receive input
           EXEC CICS RECEIVE
               INTO(WS-INPUT-AREA)
               LENGTH(WS-INPUT-LENGTH)
               RESP(WS-RESP)
           END-EXEC.
       
      * Read from file
           EXEC CICS READ
               FILE('CUSTMAST')
               INTO(CUSTOMER-RECORD)
               RIDFLD(WS-CUST-ID)
               RESP(WS-RESP)
           END-EXEC
           
           EVALUATE WS-RESP
               WHEN DFHRESP(NORMAL)
                   PERFORM PROCESS-CUSTOMER
               WHEN DFHRESP(NOTFND)
                   DISPLAY 'CUSTOMER NOT FOUND'
               WHEN OTHER
                   PERFORM ERROR-ROUTINE
           END-EVALUATE.
       
      * Write/Update
           EXEC CICS WRITE
               FILE('CUSTMAST')
               FROM(CUSTOMER-RECORD)
               RIDFLD(WS-CUST-ID)
               RESP(WS-RESP)
           END-EXEC.
           
           EXEC CICS REWRITE
               FILE('CUSTMAST')
               FROM(CUSTOMER-RECORD)
               RESP(WS-RESP)
           END-EXEC.
       
      * Link to another program
           EXEC CICS LINK
               PROGRAM('CUSTVAL')
               COMMAREA(WS-COMMAREA)
               LENGTH(LENGTH OF WS-COMMAREA)
               RESP(WS-RESP)
           END-EXEC.
       
      * Return to CICS
           EXEC CICS RETURN
           END-EXEC.
       
      * Send output
           EXEC CICS SEND
               FROM(WS-OUTPUT-AREA)
               LENGTH(WS-OUTPUT-LENGTH)
               ERASE
           END-EXEC.

Copybooks

Data Structure Copybook

      *****************************************************************
      * COPYBOOK: CUSTCOPY                                           *
      * PURPOSE:  Customer record structure                          *
      *****************************************************************
       01  CUSTOMER-RECORD.
           05  CUST-KEY.
               10  CUST-ID             PIC 9(10).
           05  CUST-PERSONAL.
               10  CUST-FIRST-NAME     PIC X(30).
               10  CUST-LAST-NAME      PIC X(30).
               10  CUST-DOB            PIC 9(08).
           05  CUST-ADDRESS.
               10  CUST-STREET         PIC X(40).
               10  CUST-CITY           PIC X(30).
               10  CUST-STATE          PIC X(02).
               10  CUST-ZIP            PIC 9(05).
           05  CUST-FINANCIAL.
               10  CUST-BALANCE        PIC S9(11)V99 COMP-3.
               10  CUST-CREDIT-LIMIT   PIC S9(09)V99 COMP-3.
           05  CUST-AUDIT.
               10  CUST-CREATE-DATE    PIC 9(08).
               10  CUST-UPDATE-DATE    PIC 9(08).
               10  CUST-UPDATE-USER    PIC X(08).

Using Copybook

       DATA DIVISION.
       FILE SECTION.
       FD  CUSTOMER-FILE.
       COPY CUSTCOPY.
       
       WORKING-STORAGE SECTION.
       01  WS-CUSTOMER-REC.
       COPY CUSTCOPY.

Subprograms & Calls

Main Program

       WORKING-STORAGE SECTION.
       01  WS-PARM-AREA.
           05  PARM-FUNCTION       PIC X(08).
           05  PARM-INPUT          PIC X(100).
           05  PARM-OUTPUT         PIC X(200).
           05  PARM-RETURN-CODE    PIC 9(04) COMP.
       
       PROCEDURE DIVISION.
           MOVE 'VALIDATE' TO PARM-FUNCTION
           MOVE CUSTOMER-ID TO PARM-INPUT
           
           CALL 'VALIDSUB' USING WS-PARM-AREA
           
           IF PARM-RETURN-CODE = 0
               PERFORM PROCESS-VALID-CUSTOMER
           ELSE
               PERFORM HANDLE-VALIDATION-ERROR
           END-IF.

Subprogram

       IDENTIFICATION DIVISION.
       PROGRAM-ID. VALIDSUB.
       
       DATA DIVISION.
       LINKAGE SECTION.
       01  LS-PARM-AREA.
           05  LS-FUNCTION         PIC X(08).
           05  LS-INPUT            PIC X(100).
           05  LS-OUTPUT           PIC X(200).
           05  LS-RETURN-CODE      PIC 9(04) COMP.
       
       PROCEDURE DIVISION USING LS-PARM-AREA.
           MOVE ZERO TO LS-RETURN-CODE
           
           EVALUATE LS-FUNCTION
               WHEN 'VALIDATE'
                   PERFORM VALIDATE-INPUT
               WHEN 'FORMAT  '
                   PERFORM FORMAT-OUTPUT
               WHEN OTHER
                   MOVE 9999 TO LS-RETURN-CODE
           END-EVALUATE
           
           GOBACK.
       
       VALIDATE-INPUT.
           IF LS-INPUT IS NUMERIC
               MOVE 'VALID INPUT' TO LS-OUTPUT
           ELSE
               MOVE 'INVALID INPUT' TO LS-OUTPUT
               MOVE 0001 TO LS-RETURN-CODE
           END-IF.

Date & Time Functions

Intrinsic Functions

       01  WS-CURRENT-DATE-TIME.
           05  WS-CURRENT-DATE.
               10  WS-CURRENT-YEAR     PIC 9(04).
               10  WS-CURRENT-MONTH    PIC 9(02).
               10  WS-CURRENT-DAY      PIC 9(02).
           05  WS-CURRENT-TIME.
               10  WS-CURRENT-HOUR     PIC 9(02).
               10  WS-CURRENT-MINUTE   PIC 9(02).
               10  WS-CURRENT-SECOND   PIC 9(02).
           05  WS-CURRENT-MS           PIC 9(02).
           05  WS-DIFF-FROM-GMT        PIC S9(04).
       
       PROCEDURE DIVISION.
      * Get current date/time
           MOVE FUNCTION CURRENT-DATE TO WS-CURRENT-DATE-TIME
           DISPLAY 'DATE: ' WS-CURRENT-DATE
           DISPLAY 'TIME: ' WS-CURRENT-TIME.
       
      * Date arithmetic
           COMPUTE WS-INTEGER-DATE =
               FUNCTION INTEGER-OF-DATE(20240115)
           ADD 30 TO WS-INTEGER-DATE
           COMPUTE WS-NEW-DATE =
               FUNCTION DATE-OF-INTEGER(WS-INTEGER-DATE).
       
      * Date validation
           IF FUNCTION TEST-DATE-YYYYMMDD(WS-DATE-FIELD) = 0
               DISPLAY 'VALID DATE'
           ELSE
               DISPLAY 'INVALID DATE'
           END-IF.

Error Handling

Comprehensive Error Handling

       01  WS-ERROR-HANDLING.
           05  WS-ERROR-OCCURRED       PIC X(01) VALUE 'N'.
               88  ERROR-YES           VALUE 'Y'.
               88  ERROR-NO            VALUE 'N'.
           05  WS-ERROR-CODE           PIC 9(04) COMP VALUE ZERO.
           05  WS-ERROR-MESSAGE        PIC X(100) VALUE SPACES.
       
       PROCEDURE DIVISION.
           PERFORM 1000-INITIALIZE
           IF ERROR-NO
               PERFORM 2000-PROCESS
           END-IF
           PERFORM 3000-FINALIZE
           
           IF ERROR-YES
               MOVE WS-ERROR-CODE TO RETURN-CODE
           END-IF
           
           STOP RUN.
       
       1000-INITIALIZE.
           OPEN INPUT MASTER-FILE
           IF NOT FILE-SUCCESS
               MOVE 'Y' TO WS-ERROR-OCCURRED
               MOVE 0100 TO WS-ERROR-CODE
               STRING 'OPEN ERROR: ' DELIMITED BY SIZE
                      WS-FILE-STATUS DELIMITED BY SIZE
                   INTO WS-ERROR-MESSAGE
               PERFORM 9000-LOG-ERROR
           END-IF.
       
       9000-LOG-ERROR.
           DISPLAY 'ERROR CODE: ' WS-ERROR-CODE
           DISPLAY 'ERROR MSG:  ' WS-ERROR-MESSAGE
           
      * Write to error log file
           WRITE ERROR-LOG-RECORD FROM WS-ERROR-HANDLING.

Compilation & Linkage

Compile JCL

//COMPILE  EXEC PGM=IGYCRCTL,
//         PARM='LIST,MAP,XREF,RENT,SSRANGE'
//STEPLIB  DD DSN=IGY.V6R3M0.SIGYCOMP,DISP=SHR
//SYSLIB   DD DSN=PROD.COPYLIB,DISP=SHR
//         DD DSN=DB2.DCLGENS,DISP=SHR
//SYSIN    DD DSN=DEV.SOURCE(CUSTINQ),DISP=SHR
//SYSPRINT DD SYSOUT=*
//SYSLIN   DD DSN=&&LOADSET,
//         DISP=(MOD,PASS),
//         SPACE=(CYL,(1,1)),
//         UNIT=SYSDA

Link-Edit JCL

//LKED     EXEC PGM=IEWL,
//         PARM='LIST,MAP,XREF,RENT,REUS'
//SYSLIB   DD DSN=CEE.SCEELKED,DISP=SHR
//         DD DSN=DB2.SDSNLOAD,DISP=SHR
//         DD DSN=CICS.SDFHLOAD,DISP=SHR
//SYSLIN   DD DSN=&&LOADSET,DISP=(OLD,DELETE)
//         DD DDNAME=SYSIN
//SYSLMOD  DD DSN=PROD.LOAD(CUSTINQ),DISP=SHR
//SYSPRINT DD SYSOUT=*
//SYSIN    DD *
  ENTRY CUSTINQ
  NAME CUSTINQ(R)
/*

Best Practices

1. Use COMP-3 for numeric calculations - Better performance 2. Avoid ALTER statement - Use structured programming 3. Use explicit scope terminators - END-IF, END-PERFORM, etc. 4. Initialize variables - Use VALUE clauses or INITIALIZE 5. Check FILE STATUS - After every I/O operation 6. Use meaningful names - Self-documenting code 7. Modularize code - Small, focused paragraphs 8. Document complex logic - Comments for maintainability 9. Handle errors comprehensively - Check SQLCODE, FILE-STATUS 10. Use copybooks - Reduce redundancy, ensure consistency

JCL & Batch Processing

JCL Fundamentals

Job Statement

//JOBNAME  JOB (account),'programmer-name',
//         CLASS=A,
//         MSGCLASS=X,
//         MSGLEVEL=(1,1),
//         NOTIFY=&SYSUID,
//         REGION=0M,
//         TIME=(0,30)

Parameters:

• CLASS: Job class for execution priority
• MSGCLASS: Output class for job logs
• MSGLEVEL: (statements,messages) for job log detail
• NOTIFY: User to notify on completion
• REGION: Memory allocation (0M = unlimited)
• TIME: Maximum execution time (minutes,seconds)

EXEC Statement

//STEPNAME EXEC PGM=program-name,
//         PARM='parameters',
//         COND=(code,operator),
//         TIME=(0,10),
//         REGION=4M

Or call a procedure:

//STEPNAME EXEC procedure-name,
//         PARM.stepname='parameters'

DD Statement

//ddname   DD DSN=dataset-name,
//         DISP=(status,normal-end,abnormal-end),
//         SPACE=(unit,(primary,secondary),RLSE),
//         UNIT=device-type,
//         DCB=(RECFM=FB,LRECL=80,BLKSIZE=800)

Common JCL Patterns

1. Copy Dataset

//COPY     EXEC PGM=IEBGENER
//SYSPRINT DD SYSOUT=*
//SYSIN    DD DUMMY
//SYSUT1   DD DSN=SOURCE.DATA,DISP=SHR
//SYSUT2   DD DSN=TARGET.DATA,
//         DISP=(NEW,CATLG,DELETE),
//         SPACE=(TRK,(100,10),RLSE),
//         UNIT=SYSDA,
//         DCB=(RECFM=FB,LRECL=80,BLKSIZE=27920)

2. Sort Data

//SORT     EXEC PGM=SORT
//SYSOUT   DD SYSOUT=*
//SORTIN   DD DSN=INPUT.DATA,DISP=SHR
//SORTOUT  DD DSN=SORTED.DATA,
//         DISP=(NEW,CATLG,DELETE),
//         SPACE=(CYL,(10,5),RLSE),
//         UNIT=SYSDA
//SYSIN    DD *
  SORT FIELDS=(1,10,CH,A,20,5,ZD,D)
  SUM FIELDS=(30,8,ZD)
/*

3. Conditional Execution

//STEP1    EXEC PGM=PROG1
//STEP2    EXEC PGM=PROG2,COND=(0,NE,STEP1)
//STEP3    EXEC PGM=PROG3,COND=((0,NE,STEP1),(0,NE,STEP2))

Or using IF/THEN/ELSE:

//STEP1    EXEC PGM=PROG1
//IF1      IF (STEP1.RC = 0) THEN
//STEP2      EXEC PGM=PROG2
//ENDIF    ENDIF

4. GDG (Generation Data Group)

//DEFINE   EXEC PGM=IDCAMS
//SYSPRINT DD SYSOUT=*
//SYSIN    DD *
  DEFINE GDG(NAME(MY.GDG.BASE) -
    LIMIT(7) -
    SCRATCH -
    NOEMPTY)
/*

//CREATE   EXEC PGM=IEBGENER
//SYSUT1   DD *
data
/*
//SYSUT2   DD DSN=MY.GDG.BASE(+1),
//         DISP=(NEW,CATLG,DELETE),
//         SPACE=(TRK,(10,5)),
//         UNIT=SYSDA
//SYSIN    DD DUMMY
//SYSPRINT DD SYSOUT=*

Reference GDG versions:

• (0) = Current generation
• (+1) = New generation
• (-1) = Previous generation
• (-2) = Two generations back

5. Concatenated Datasets

//INPUT    DD DSN=FILE1.DATA,DISP=SHR
//         DD DSN=FILE2.DATA,DISP=SHR
//         DD DSN=FILE3.DATA,DISP=SHR

6. In-stream Data

//SYSIN    DD *
input data line 1
input data line 2
/*

//Or with delimiter:
//SYSIN    DD DATA,DLM=$$
input data with /* inside
$$

7. Temporary Datasets

//TEMP1    DD DSN=&&TEMPFILE,
//         DISP=(NEW,PASS),
//         SPACE=(CYL,(5,1)),
//         UNIT=SYSDA

JCL Utilities

IEFBR14 - Allocate/Delete

//ALLOC    EXEC PGM=IEFBR14
//NEWFILE  DD DSN=MY.NEW.FILE,
//         DISP=(NEW,CATLG,DELETE),
//         SPACE=(CYL,(10,5)),
//         UNIT=SYSDA,
//         DCB=(RECFM=FB,LRECL=80,BLKSIZE=27920)

//DELETE   EXEC PGM=IEFBR14
//OLDFILE  DD DSN=MY.OLD.FILE,DISP=(OLD,DELETE,DELETE)

IEBGENER - Copy

//COPY     EXEC PGM=IEBGENER
//SYSPRINT DD SYSOUT=*
//SYSUT1   DD DSN=INPUT,DISP=SHR
//SYSUT2   DD DSN=OUTPUT,DISP=(NEW,CATLG)
//SYSIN    DD *
  GENERATE MAXFLDS=1
  RECORD FIELD=(80,1,,1)
/*

IEBCOPY - Copy PDS

//COPY     EXEC PGM=IEBCOPY
//SYSPRINT DD SYSOUT=*
//INDD     DD DSN=SOURCE.PDS,DISP=SHR
//OUTDD    DD DSN=TARGET.PDS,DISP=(NEW,CATLG)
//SYSIN    DD *
  COPY INDD=INDD,OUTDD=OUTDD
  SELECT MEMBER=(MEM1,MEM2,MEM3)
/*

IEBUPDTE - Update PDS

//UPDATE   EXEC PGM=IEBUPDTE
//SYSPRINT DD SYSOUT=*
//SYSUT1   DD DSN=MY.PDS,DISP=OLD
//SYSUT2   DD DSN=MY.PDS,DISP=OLD
//SYSIN    DD *
./ ADD NAME=MEMBER1,LIST=ALL
source line 1
source line 2
./ ENDUP
/*

DFSORT - Sort/Merge

//SORT     EXEC PGM=SORT
//SYSOUT   DD SYSOUT=*
//SORTIN   DD DSN=INPUT,DISP=SHR
//SORTOUT  DD DSN=OUTPUT,DISP=(NEW,CATLG)
//SYSIN    DD *
  SORT FIELDS=(1,10,CH,A,15,5,ZD,D)
  INCLUDE COND=(20,2,CH,EQ,C'NY')
  OUTREC FIELDS=(1,80,100X)
/*

Advanced sort:

//SYSIN    DD *
  SORT FIELDS=(1,10,CH,A)
  SUM FIELDS=(20,8,ZD,30,8,PD)
  OUTFIL FNAMES=OUT1,INCLUDE=(15,2,CH,EQ,C'CA'),
    BUILD=(1,10,15,20)
  OUTFIL FNAMES=OUT2,INCLUDE=(15,2,CH,EQ,C'NY'),
    BUILD=(1,10,15,20)
/*

IDCAMS - Catalog Management

//IDCAMS   EXEC PGM=IDCAMS
//SYSPRINT DD SYSOUT=*
//SYSIN    DD *
  /* Define VSAM cluster */
  DEFINE CLUSTER( -
    NAME(MY.VSAM.FILE) -
    KEYS(10 0) -
    RECORDSIZE(100 100) -
    CYLINDERS(5 1) -
    INDEXED -
    SHAREOPTIONS(2 3)) -
  DATA( -
    NAME(MY.VSAM.FILE.DATA)) -
  INDEX( -
    NAME(MY.VSAM.FILE.INDEX))
    
  /* Copy */
  REPRO INFILE(INDD) OUTFILE(OUTDD)
  
  /* Delete */
  DELETE MY.OLD.FILE CLUSTER
  
  /* List catalog */
  LISTCAT ENTRIES(MY.VSAM.FILE) ALL
/*

Procedures (PROCs)

Define Procedure

//MYPROC   PROC MBR=,LIB=USER.LOAD
//STEP1    EXEC PGM=PROG1
//STEPLIB  DD DSN=&LIB,DISP=SHR
//INPUT    DD DSN=INPUT.&MBR,DISP=SHR
//OUTPUT   DD DSN=OUTPUT.&MBR,
//         DISP=(NEW,CATLG,DELETE),
//         SPACE=(CYL,(5,1)),
//         UNIT=SYSDA

Call Procedure

//CALL     EXEC MYPROC,MBR=TEST01,LIB=PROD.LOAD

Override DD in Procedure

//CALL     EXEC MYPROC
//STEP1.INPUT DD DSN=ALTERNATE.INPUT,DISP=SHR
//STEP1.OUTPUT DD DSN=ALTERNATE.OUTPUT,DISP=(NEW,CATLG)

Symbolic Parameters

JCL Symbols

//SET1     SET ENV='PROD'
//SET2     SET DATE='&YYMMDD'
//STEP1    EXEC PGM=PROG1
//INPUT    DD DSN=&ENV..DATA.&DATE,DISP=SHR

System symbols:

• &SYSUID - User ID
• &SYSDATE - Current date (yyddd)
• &SYSTIME - Current time (hhmm)
• &LDAY - Day of month (dd)
• &LMONTH - Month (mm)
• &LYEAR - Year (yyyy)

Example with Symbols

//BACKUP   JOB
//SET1     SET DATE='&LYEAR&LMONTH&LDAY'
//STEP1    EXEC PGM=IEBGENER
//SYSUT1   DD DSN=PROD.DATA,DISP=SHR
//SYSUT2   DD DSN=BACKUP.DATA.D&DATE,
//         DISP=(NEW,CATLG)
//SYSIN    DD DUMMY
//SYSPRINT DD SYSOUT=*

Batch Scheduling

Job Dependencies

Using JES2 job dependency:

//JOB1     JOB
//STEP1    EXEC PGM=PROG1
//STEP2    EXEC PGM=PROG2

//JOB2     JOB
//*AFTER JOB1
//STEP1    EXEC PGM=PROG3

Using automation tools:

• TWS (Tivoli Workload Scheduler)
• Control-M
• CA-7
• Zeke

Job Submission Automation

Submit via TSO SUBMIT:

SUBMIT 'MY.JCL(JOBNAME)'

Submit via automation:

/* REXX script */
QUEUE "//JOBNAME  JOB"
QUEUE "//STEP1    EXEC PGM=IEFBR14"
"SUBMIT * END(/*)"

Performance Optimization

Efficient JCL Practices

1. Block size optimization:

//FILE     DD DSN=MY.DATA,
//         DCB=(RECFM=FB,LRECL=80,BLKSIZE=27920)
/* Optimal: BLKSIZE = track size - control bytes */
/* For 3390: ~27920 for LRECL=80 */

1. Space allocation:

//FILE     DD SPACE=(CYL,(primary,secondary),RLSE)
/* Use RLSE to release unused space */
/* Allocate enough primary to avoid secondary extends */

1. Temporary datasets:

//TEMP     DD DSN=&&TEMP,
//         DISP=(NEW,PASS),
//         SPACE=(CYL,(10,5)),
//         UNIT=VIO  /* Virtual I/O in memory */

1. Concatenation order:

/* Put most frequently accessed first */
//STEPLIB  DD DSN=FREQUENT.LOAD,DISP=SHR
//         DD DSN=LESS.FREQUENT.LOAD,DISP=SHR

Parallel Execution

Independent steps:

//STEP1A   EXEC PGM=PROG1A
//STEP1B   EXEC PGM=PROG1B
//STEP2    EXEC PGM=PROG2,COND=((0,NE,STEP1A),(0,NE,STEP1B))

Multiple jobs: Submit multiple independent jobs to run in parallel.

Error Handling

Return Codes

Check in JCL:

//STEP1    EXEC PGM=PROG1
//IF1      IF (STEP1.RC = 0) THEN
//SUCCESS    EXEC PGM=SUCCESS
//ELSE     ELSE
//FAILURE    EXEC PGM=FAILURE
//ENDIF    ENDIF

Condition codes:

//STEP2    EXEC PGM=PROG2,COND=(4,LT,STEP1)
/* Execute STEP2 if STEP1 RC < 4 */

/* Operators: EQ NE LT LE GT GE */

ABEND Handling

Catch ABEND:

//STEP1    EXEC PGM=PROG1
//IF1      IF (STEP1.ABEND = TRUE) THEN
//CLEANUP    EXEC PGM=CLEANUP
//ENDIF    ENDIF

Common ABENDs:

• S0C7: Data exception (invalid numeric)
• S0C4: Protection exception (memory access)
• S806: Program not found
• S322: Time out
• S013: OPEN error (dataset)

Advanced Techniques

Dynamic Allocation

IEFBR14 with dynamic:

//ALLOC    EXEC PGM=IKJEFT01
//SYSTSPRT DD SYSOUT=*
//SYSTSIN  DD *
  ALLOC DA('MY.NEW.FILE') NEW -
    SPACE(5,1) CYLINDERS -
    RECFM(F B) LRECL(80) BLKSIZE(27920) -
    UNIT(SYSDA) CATALOG
  FREE DA('MY.OLD.FILE') DELETE
/*

JCL Include

Member in JCLLIB:

//MYJOB    JOB
//         JCLLIB ORDER=MY.PROCLIB
//         INCLUDE MEMBER=COMMON
//STEP1    EXEC PGM=PROG1

COMMON member:

//STEPLIB  DD DSN=MY.LOAD.LIB,DISP=SHR
//SYSPRINT DD SYSOUT=*

REXX Integration

//REXX     EXEC PGM=IKJEFT01
//SYSEXEC  DD DSN=MY.REXX.LIB,DISP=SHR
//SYSTSPRT DD SYSOUT=*
//SYSTSIN  DD *
  %MYSCRIPT parm1 parm2
/*

Best Practices

Naming Conventions

• Job names: 8 characters max, alphanumeric
• Step names: meaningful, describe function
• DD names: standard names (SYSOUT, SYSPRINT, etc.)
• Dataset names: hierarchical, logical grouping

Documentation

//*********************************************************
//* Job: MONTHEND                                        *
//* Purpose: Monthly closing batch                       *
//* Schedule: Last day of month, 8 PM                    *
//* Dependencies: DAILYUPD must complete first           *
//* Runtime: Approximately 2 hours                       *
//* Contact: Operations Team (x1234)                     *
//*********************************************************

Testing

1. Test with small datasets first 2. Use TYPRUN=SCAN to check syntax 3. Validate with COND=(0,EQ) for conditional testing 4. Keep test and production jobs separate

Maintenance

1. Use procedures for repeated logic 2. Symbolic parameters for flexibility 3. Comment complex logic 4. Version control JCL members 5. Regular cleanup of old jobs

Troubleshooting

JCL Errors

JCL ERROR:

• Syntax errors in JCL statements
• Check JESMSGLG for details

S806 (Program not found):

//STEPLIB  DD DSN=CORRECT.LOAD.LIB,DISP=SHR

S013 (OPEN error):

• Dataset not found
• DISP conflict
• Volume not mounted
• Insufficient space

S0C4 (Protection):

• Program error accessing memory
• Check SYSUDUMP/SYSABEND

Performance Issues

Long run time:

1. Check sort efficiency 2. Optimize I/O 3. Review dataset allocation 4. Check for contention

Space issues:

/* Use RLSE to return unused space */
//FILE     DD SPACE=(CYL,(10,5),RLSE)

/* Or use compression */
//TAPE     DD DSN=BACKUP,UNIT=TAPE,
//         DISP=(NEW,CATLG),
//         DCB=TRTCH=COMP

Migration to Modern Batch

Spring Batch Equivalent

JCL batch job:

//DAILY    JOB
//STEP1    EXEC PGM=EXTRACT
//STEP2    EXEC PGM=TRANSFORM
//STEP3    EXEC PGM=LOAD

Spring Batch:

@Bean
public Job dailyJob() {
    return jobBuilderFactory.get("dailyJob")
        .start(extractStep())
        .next(transformStep())
        .next(loadStep())
        .build();
}

Airflow DAG Equivalent

from airflow import DAG
from airflow.operators.bash import BashOperator

dag = DAG('daily_batch', schedule_interval='@daily')

extract = BashOperator(task_id='extract', bash_command='extract.sh', dag=dag)
transform = BashOperator(task_id='transform', bash_command='transform.sh', dag=dag)
load = BashOperator(task_id='load', bash_command='load.sh', dag=dag)

extract >> transform >> load

Mainframe Modernization

Assessment Framework

Discovery Phase

1. Application Inventory

Component Analysis:
- Number of programs (COBOL, PL/I, Assembler, REXX)
- Total lines of code by language
- JCL jobs and procedures
- CICS/IMS transactions
- Copybooks and include files
- Database schemas (DB2, IMS, VSAM)
- Screens/maps (BMS, MFS)
- External interfaces

2. Technical Complexity Scoring

For each program, score:
- Lines of code (1-5 points based on size)
- External calls (1 point per CALL)
- File I/O operations (2 points per file)
- Database operations (2 points per table accessed)
- CICS commands (3 points if CICS-enabled)
- Dynamic program calls (5 points each)
- Computed GO TO / ALTER (10 points each - high risk)
- Embedded assembler (10 points)

Complexity Rating:
0-20: Low
21-50: Medium
51-100: High
100+: Very High

3. Business Criticality Matrix

Rate each application:
- Transaction volume (1-5)
- Revenue impact (1-5)
- Customer-facing (1-5)
- Regulatory requirements (1-5)
- Frequency of change (1-5)

Priority = Sum of ratings
20-25: Critical
15-19: High
10-14: Medium
5-9: Low

Dependency Analysis

Program Call Hierarchy:

# Extract CALL statements from COBOL
grep -r "CALL '" *.cbl | \
  sed "s/.*CALL '\([^']*\)'.*/\1/" | \
  sort | uniq > call-hierarchy.txt

# Visualize with GraphViz
echo "digraph calls {" > calls.dot
while read line; do
  echo "  \"$program\" -> \"$line\";" >> calls.dot
done < call-hierarchy.txt
echo "}" >> calls.dot
dot -Tpng calls.dot -o call-hierarchy.png

Data Dependencies:

Map:
- VSAM files → Programs that read/write
- DB2 tables → Programs that access
- Copybooks → Programs that include them
- JCL datasets → Jobs that use them
- GDG relationships

Integration Points:

Identify:
- MQ message flows
- CICS/IMS communication
- File transfers (NDM, FTP)
- Web services (z/OS Connect)
- Batch scheduling dependencies
- External system interfaces

Migration Strategies

1. Rehost (Lift & Shift)

Approach:

• Migrate to mainframe emulator (Micro Focus, LzLabs, Raincode)
• Run on x86 servers or cloud
• Minimal code changes

Pros:

• Fastest migration (6-12 months)
• Lowest risk
• Preserve existing skills
• Quick cloud benefits

Cons:

• Technical debt remains
• Limited modernization
• Ongoing emulator licensing
• Performance may vary

Best For:

• Time-sensitive migrations
• Limited budget
• Stable applications
• Skills shortage

Implementation Steps:

1. Select emulator platform
2. Install and configure emulator
3. Migrate source code
4. Recompile programs
5. Migrate data files
6. Test functionality
7. Performance tune
8. Cutover

Example: Micro Focus Enterprise Server

# Compile COBOL to native code
cblc -C "anim" CUSTINQ.cbl

# Deploy to server
cobrun CUSTINQ

# Or compile to .NET/JVM
cblc -C "netmf" CUSTINQ.cbl

2. Replatform

Approach:

• Migrate to modern platform
• Keep similar architecture
• Update technology stack

Pros:

• Moderate risk
• Some modernization
• Better performance
• Cloud-native deployment

Cons:

• Longer timeline (12-24 months)
• More complex
• Requires reskilling
• Testing overhead

Best For:

• Moderate complexity apps
• Need for scalability
• Hybrid approach
• Gradual modernization

Implementation: COBOL to Java

COBOL Source:

       PROCEDURE DIVISION.
           PERFORM READ-CUSTOMER
           IF CUSTOMER-FOUND
               PERFORM CALCULATE-BALANCE
               PERFORM UPDATE-RECORD
           END-IF.

Java Equivalent:

public class CustomerProcessor {
    public void processCustomer(String customerId) {
        Customer customer = customerRepository.findById(customerId);
        if (customer != null) {
            BigDecimal balance = calculateBalance(customer);
            customer.setBalance(balance);
            customerRepository.save(customer);
        }
    }
    
    private BigDecimal calculateBalance(Customer customer) {
        return customer.getCredits()
            .subtract(customer.getDebits());
    }
}

3. Refactor (Microservices)

Approach:

• Decompose monolith
• Build microservices
• API-first design

Pros:

• Full modernization
• Cloud-native
• Agile development
• Technology flexibility

Cons:

• Longest timeline (24-36 months)
• Highest risk
• Complete reskilling
• Significant investment

Best For:

• Complex applications
• Long-term strategy
• Digital transformation
• Scalability needs

Architecture Transformation:

Before (Monolithic):

CICS Region
├── Customer Management Programs
├── Account Management Programs
├── Transaction Processing Programs
└── Reporting Programs

After (Microservices):

API Gateway
├── Customer Service (Spring Boot)
├── Account Service (Node.js)
├── Transaction Service (Java)
└── Reporting Service (Python)

Example Microservice:

@RestController
@RequestMapping("/api/v1/customers")
public class CustomerController {
    
    @Autowired
    private CustomerService customerService;
    
    @GetMapping("/{id}")
    public ResponseEntity<CustomerDTO> getCustomer(
            @PathVariable String id) {
        return customerService.findById(id)
            .map(ResponseEntity::ok)
            .orElse(ResponseEntity.notFound().build());
    }
    
    @PostMapping
    public ResponseEntity<CustomerDTO> createCustomer(
            @Valid @RequestBody CustomerDTO customer) {
        CustomerDTO created = customerService.create(customer);
        return ResponseEntity.created(
            URI.create("/api/v1/customers/" + created.getId()))
            .body(created);
    }
    
    @PutMapping("/{id}")
    public ResponseEntity<CustomerDTO> updateCustomer(
            @PathVariable String id,
            @Valid @RequestBody CustomerDTO customer) {
        return customerService.update(id, customer)
            .map(ResponseEntity::ok)
            .orElse(ResponseEntity.notFound().build());
    }
}

4. Retire & Replace

Approach:

• Replace with COTS/SaaS
• Commercial alternatives
• Modern platforms

Pros:

• No development
• Modern features
• Vendor support
• Quick deployment

Cons:

• Customization limits
• Data migration
• Process changes
• Vendor lock-in

Best For:

• Commodity functions
• Standard processes
• Budget constraints
• Quick wins

Hybrid Integration Patterns

Pattern 1: API Wrapper (Strangler Fig)

Expose mainframe as API:

External Systems → API Gateway → z/OS Connect → CICS/IMS

z/OS Connect Configuration:

{
  "serviceName": "CustomerInquiry",
  "serviceProvider": "CICS",
  "endpoint": {
    "program": "CUSTINQ",
    "commarea": {
      "customerNumber": "string",
      "requestType": "string"
    }
  },
  "apiDefinition": {
    "path": "/customers/{id}",
    "method": "GET",
    "requestMapping": {
      "id": "customerNumber"
    }
  }
}

API Gateway (Kong/Apigee):

services:
  - name: customer-service
    url: https://zosconnect.company.com/api/customers
    routes:
      - name: get-customer
        paths:
          - /api/customers
    plugins:
      - name: rate-limiting
        config:
          minute: 100
      - name: authentication
        config:
          type: oauth2

Pattern 2: Data Replication (CDC)

Real-time data sync:

Mainframe DB2 → CDC (IBM InfoSphere) → Kafka → Cloud Database

CDC Configuration:

-- Source (DB2 z/OS)
CREATE TABLE CUSTOMER (
    CUST_ID INT PRIMARY KEY,
    CUST_NAME VARCHAR(50),
    LAST_UPDATE TIMESTAMP
);

-- Enable CDC
ALTER TABLE CUSTOMER 
    DATA CAPTURE CHANGES;

-- CDC capture
CREATE SUBSCRIPTION CUST_SUB FOR TABLE CUSTOMER;

Kafka Consumer:

@KafkaListener(topics = "mainframe.customer")
public void processCustomerUpdate(CustomerEvent event) {
    switch(event.getOperation()) {
        case INSERT:
            cloudRepository.save(event.getData());
            break;
        case UPDATE:
            cloudRepository.update(event.getData());
            break;
        case DELETE:
            cloudRepository.delete(event.getKey());
            break;
    }
}

Pattern 3: Event-Driven Integration

Asynchronous messaging:

CICS/IMS → IBM MQ → Message Broker → Cloud Services

CICS to MQ:

       EXEC CICS WRITEQ TS
           QUEUE('CUSTOMER.UPDATES')
           FROM(CUSTOMER-RECORD)
           LENGTH(LENGTH OF CUSTOMER-RECORD)
       END-EXEC.

Spring Boot Consumer:

@JmsListener(destination = "CUSTOMER.UPDATES")
public void handleCustomerUpdate(CustomerMessage msg) {
    log.info("Received customer update: {}", msg.getCustomerId());
    customerService.processUpdate(msg);
}

Data Migration Strategy

Phase 1: Analysis

1. Data profiling
   - Volume
   - Quality issues
   - Relationships
   - Dependencies

2. Schema mapping
   - Source to target
   - Type conversions
   - Transformation rules
   - Default values

3. Validation rules
   - Business rules
   - Referential integrity
   - Data quality checks

Phase 2: Design

VSAM to Relational Database:

Source (VSAM):

01  CUSTOMER-RECORD.
    05  CUST-ID             PIC 9(10).
    05  CUST-NAME           PIC X(50).
    05  CUST-BALANCE        PIC S9(11)V99 COMP-3.
    05  CUST-STATUS         PIC X(01).

Target (PostgreSQL):

CREATE TABLE customer (
    cust_id BIGINT PRIMARY KEY,
    cust_name VARCHAR(50) NOT NULL,
    cust_balance NUMERIC(13,2) DEFAULT 0.00,
    cust_status CHAR(1) CHECK (cust_status IN ('A','I')),
    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    updated_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
    CONSTRAINT valid_balance CHECK (cust_balance >= 0)
);

CREATE INDEX idx_customer_status ON customer(cust_status);
CREATE INDEX idx_customer_name ON customer(cust_name);

Phase 3: ETL Process

Extract from DB2:

-- DB2 z/OS
UNLOAD EXTERNAL TABLE CUSTOMER
TO '/data/customer_extract.dat'
USING DELIMITER ','
WITH HEADER;

Transform & Load:

import pandas as pd
from sqlalchemy import create_engine

# Extract
df = pd.read_csv('customer_extract.dat')

# Transform
df['cust_id'] = pd.to_numeric(df['cust_id'])
df['cust_balance'] = pd.to_numeric(df['cust_balance']) / 100
df['cust_status'] = df['cust_status'].str.strip()
df['created_at'] = pd.to_datetime('now')
df['updated_at'] = pd.to_datetime('now')

# Validate
assert df['cust_id'].is_unique, "Duplicate customer IDs"
assert df['cust_status'].isin(['A', 'I']).all(), "Invalid status"
assert (df['cust_balance'] >= 0).all(), "Negative balance"

# Load
engine = create_engine('postgresql://host/database')
df.to_sql('customer', engine, if_exists='append', index=False)

print(f"Loaded {len(df)} records")

Phase 4: Validation

Reconciliation Query:

-- Source count
SELECT COUNT(*) FROM CUSTOMER WHERE CUST_STATUS = 'A';

-- Target count
SELECT COUNT(*) FROM customer WHERE cust_status = 'A';

-- Balance total
SELECT SUM(CUST_BALANCE) FROM CUSTOMER;
SELECT SUM(cust_balance) FROM customer;

DevOps Integration

CI/CD Pipeline for Mainframe

1. Source Control (Git):

# Clone mainframe source
git clone https://github.com/company/mainframe-apps.git

# Structure
mainframe-apps/
├── cobol/
│   ├── programs/
│   ├── copybooks/
│   └── tests/
├── jcl/
├── bms/
├── db2/
└── .github/workflows/

2. Build Pipeline (Jenkins):

pipeline {
    agent any
    
    stages {
        stage('Checkout') {
            steps {
                git 'https://github.com/company/mainframe-apps.git'
            }
        }
        
        stage('Compile') {
            steps {
                script {
                    // Use Zowe CLI or DBB
                    sh '''
                        zowe files upload file-to-data-set \
                            "cobol/programs/CUSTINQ.cbl" \
                            "DEV.SOURCE(CUSTINQ)"
                        
                        zowe jobs submit data-set "DEV.JCL(COMPILE)" \
                            --wait-for-output
                    '''
                }
            }
        }
        
        stage('Test') {
            steps {
                sh '''
                    zowe jobs submit data-set "DEV.JCL(UNITTEST)" \
                        --wait-for-output
                '''
            }
        }
        
        stage('Deploy') {
            when {
                branch 'main'
            }
            steps {
                sh '''
                    zowe files copy data-set \
                        "DEV.LOAD(CUSTINQ)" \
                        "PROD.LOAD(CUSTINQ)"
                '''
            }
        }
    }
}

3. Testing Strategy:

// Unit test for migrated code
@Test
public void testCalculateBalance() {
    Customer customer = new Customer();
    customer.setCredits(new BigDecimal("1000.00"));
    customer.setDebits(new BigDecimal("250.50"));
    
    BigDecimal balance = customerService.calculateBalance(customer);
    
    assertEquals(new BigDecimal("749.50"), balance);
}

// Integration test
@SpringBootTest
@AutoConfigureMockMvc
public class CustomerAPITest {
    @Autowired
    private MockMvc mockMvc;
    
    @Test
    public void testGetCustomer() throws Exception {
        mockMvc.perform(get("/api/customers/1234567890"))
            .andExpect(status().isOk())
            .andExpect(jsonPath("$.customerId").value("1234567890"))
            .andExpect(jsonPath("$.balance").value(1000.00));
    }
}

Migration Tools

Automated Code Conversion

1. COBOL to Java (AWS Blu Age):

# Convert COBOL program
blu-age convert \
    --source cobol/CUSTINQ.cbl \
    --target java/CustomerInquiry.java \
    --framework spring-boot

# Generated Spring Boot service

2. Micro Focus Visual COBOL:

# Compile COBOL to .NET
cobc -m -x CUSTINQ.cbl -o CustomerInquiry.dll

# Or to JVM
cobc -m -x CUSTINQ.cbl -t java -o CustomerInquiry.jar

3. LzLabs SDM (Software Defined Mainframe):

# Deploy workload to container
lzlabs deploy \
    --programs cobol/*.cbl \
    --jcl jcl/*.jcl \
    --data vsam/*.dat \
    --target kubernetes

Static Analysis

Analyze code quality:

# SonarQube for COBOL
sonar-scanner \
    -Dsonar.projectKey=mainframe-app \
    -Dsonar.sources=cobol \
    -Dsonar.language=cobol

# Code metrics
sloccount cobol/
# Output: Lines of code, complexity, maintainability

Migration Timeline Example

18-Month Replatform Project

Months 1-3: Discovery & Planning

• Application inventory
• Dependency mapping
• Complexity assessment
• Tool selection
• Team training

Months 4-6: Proof of Concept

• Select pilot application
• Convert and deploy
• Performance testing
• Lessons learned

Months 7-12: Bulk Migration

• Wave 1: Low complexity (months 7-8)
• Wave 2: Medium complexity (months 9-10)
• Wave 3: High complexity (months 11-12)

Months 13-15: Integration & Testing

• End-to-end testing
• Performance tuning
• Security hardening
• User acceptance testing

Months 16-18: Cutover & Stabilization

• Production deployment
• Parallel running
• Issue resolution
• Knowledge transfer

Post-Migration

Decommissioning Checklist

□ Verify all data migrated
□ Confirm all integrations working
□ Archive mainframe code
□ Document new architecture
□ Train operations team
□ Establish monitoring
□ Update disaster recovery
□ Release mainframe resources
□ Contract negotiations
□ Celebrate success!

Cost Optimization

Before (Mainframe):
- MIPS charges: $1M/year
- Storage: $200K/year
- Software licenses: $500K/year
Total: $1.7M/year

After (Cloud):
- Compute (VMs): $300K/year
- Storage: $50K/year
- Software licenses: $200K/year
Total: $550K/year

Savings: $1.15M/year (68% reduction)
ROI: Migration cost recovered in 18 months

Success Factors

1. Executive sponsorship - Critical for funding and support 2. Incremental approach - Reduce risk, show progress 3. Skill development - Train team on new technologies 4. Automated testing - Ensure functional equivalence 5. Data quality - Clean data before migration 6. Documentation - Knowledge capture throughout 7. Change management - Prepare organization 8. Risk mitigation - Backout plans at every stage

z/OS System Administration

System Operations

IPL (Initial Program Load)

Cold start:

1. Issue QUIESCE command
2. Select IPL volume
3. Specify LOADPARM parameters
4. Monitor console for IPL messages

Warm start (system restart):

C U=CPU,RESTART        /* Restart specific CPU */

Console Commands

Display commands:

/D A,ALL                /* All active address spaces */
/D A,L                  /* Address spaces by LPAR */
/D M=CPU               /* CPU usage */
/D M=STOR              /* Storage usage */
/D U,ALL               /* All units */
/D PROG,APF            /* APF authorized libraries */
/D PROD,STATE          /* Product status */
/D IOS,CONFIG          /* I/O configuration */
/D SMS,STORGRP(ALL)    /* Storage groups */

System control:

/V OFFLINE             /* Vary device offline */
/V ONLINE              /* Vary device online */
/STOP subsystem        /* Stop subsystem */
/START subsystem       /* Start subsystem */
/CANCEL jobname        /* Cancel job */
/F jobname,command     /* Send command to job */

JES2 Commands

Job management:

/$DA                   /* Display active jobs */
/$DQ                   /* Display job queue */
/$DJ jobname           /* Display job status */
/$PJOBNAME            /* Purge job */
/$CANCEL jobname       /* Cancel job */

Device management:

/$DSPL                 /* Display spool */
/$DDEV                 /* Display devices */
/$S PRINTER1           /* Start printer */
/$P PRINTER1           /* Stop printer */

Configuration:

/$T A=ALL              /* Display all options */
/$T JOBCLASS(A),QHOLD  /* Hold job class */
/$T JOBCLASS(A),QREL   /* Release job class */

Storage Management

SMS (Storage Management Subsystem)

Storage groups:

DEFINE STORAGEGROUP(STGRP1)
  VOLUMES(VOL001 VOL002 VOL003)
  VSAMDATACLASS(VSAM1)
  AVGREC(U)
  
LISTCAT STORAGEGROUP(STGRP1) ALL

Management classes:

DEFINE MANAGEMENTCLASS(MGMT1)
  MIGRATION(DAYS(30))
  BACKUP(DAYS(7))
  RETENTION(DAYS(365))

Data classes:

DEFINE DATACLASS(DCLASS1)
  RECFM(FB)
  LRECL(80)
  SPACEAVG(100 CYL)

Dataset Migration & Backup

DFHSM (HSM) commands:

HSEND MIGRATE DSN('USER.DATA.**')
HSEND RECALL DSN('USER.DATA.FILE')
HSEND BACKUP DSN('USER.DATA.**')
HSEND RECOVER DSN('USER.DATA.FILE')
HSEND DELETE BACKUP DSN('USER.DATA.FILE')

ADRDSSU (DSS) backup:

//BACKUP   EXEC PGM=ADRDSSU
//SYSPRINT DD SYSOUT=*
//DASD1    DD UNIT=SYSDA,VOL=SER=WORK01,DISP=OLD
//TAPE1    DD DSN=BACKUP.TAPE,UNIT=TAPE,
//            VOL=SER=TAPE01,DISP=(NEW,CATLG)
//SYSIN    DD *
  DUMP DATASET( -
    INCLUDE(**) -
    BY(DSORG(PS,PO)) ) -
    INDDNAME(DASD1) -
    OUTDDNAME(TAPE1) -
    COMPRESS
/*

System Maintenance

SMP/E (System Modification Program)

Apply maintenance:

SET BDY(GLOBAL)
RECEIVE S(PTF12345)
SET BDY(TARGET)
APPLY S(PTF12345) CHECK
APPLY S(PTF12345)
SET BDY(DLIB)
ACCEPT S(PTF12345)

List maintenance:

LIST PTFS
LIST SYSMOD(PTF12345)
LIST MOD(IEFBR14)

LLA (Library Lookaside)

Refresh LLA:

/F LLA,UPDATE=xx
/F LLA,REFRESH
/F LLA,DISPLAY

System Parameters (PARMLIB)

IPL parameters (LOADxx):

SYSCAT    CATALOG.MASTER
NUCLEUS   01
PARMLIB   SYS1.PARMLIB
SYSPLEX   PLEXNAME

System options (IEASYSxx):

APF=00           /* APF list */
CMD=00           /* Commands */
CON=00           /* Consoles */
GRS=YES          /* Global Resource Serialization */
PROG=00          /* Programs */
SMF=00           /* SMF parameters */

Subsystem Management

Subsystem Initialization Table (IEFSSNxx)

SUBSYS SUBNAME(JES2)
  INITRTN(HASJES20)
  INITPARM('WARM')
  
SUBSYS SUBNAME(CICS)
  INITRTN(DFHSIP)
  
SUBSYS SUBNAME(DB2P)
  INITRTN(DSNMSTR)

Dynamic Subsystem Control

Add subsystem:

/SETSSI ADD,S=subsys,I=initrtn

Delete subsystem:

/SETSSI DELETE,S=subsys

Performance Monitoring

RMF (Resource Measurement Facility)

Start RMF:

/S RMF,PARM=xx
/F RMF,START(type)

RMF reports:

• Monitor I: CPU activity
• Monitor II: DASD activity
• Monitor III: Workload activity
• OVERVIEW: System summary
• WKLD: Workload by service class

SMF (System Management Facilities)

SMF parameters (SMFPRMxx):

SYS(INTERVAL(30))
SUB(TYPE(0,30,70,80,89,90))
REC(EXITS(SYS(ON),SUBSYS(ON)))
JWT(INTERVAL(15))

Extract SMF data:

//EXTRACT  EXEC PGM=IFASMFDP
//SYSPRINT DD SYSOUT=*
//DUMPIN   DD DISP=SHR,DSN=SYS1.MANX
//DUMPOUT  DD DISP=(NEW,CATLG),DSN=SMF.EXTRACT,
//            SPACE=(CYL,(50,10))
//SYSIN    DD *
  INDD(DUMPIN,OPTIONS(DUMP))
  OUTDD(DUMPOUT,TYPE(30,80))
  DATE(2024001,2024365)
/*

Automation

Automation Table (ATx00)

Define automated responses:

IF MSGID = 'IEF404I' THEN
  EXEC(PGM=CLEANUP,PARM='&JOBNAME')
ENDIF

IF TEXT = 'HASP395' THEN
  DO
    SUB('RESTART.JCL')
  DONE
ENDIF

NetView Automation

AutoOps definitions:

IF (MSGID = 'IEE361I') THEN
  DO
    WRITE('MOUNT VOLUME &1')
    WAIT(MSGID = 'IEA511I')
    CLIST(AUTOVOLM,&1)
  DONE
ENDIF

Disaster Recovery

Parallel Sysplex

Coupling Facility structures:

STRNAME  TYPE    SIZE     POLICY
LOCK01   LOCK    256KB    SIZE(256)
SCA01    SCA     512KB    SIZE(512)
LOGR01   LOGR    1024KB   SIZE(1024)

XCF (Cross-System Coupling Facility):

/D XCF                  /* Display XCF status */
/D XCF,STR,STRNAME=ALL  /* Display structures */
/D XCF,GROUP            /* Display groups */

GDPS (Geographically Dispersed Parallel Sysplex)

Configuration:

• Primary site with active workload
• Secondary site with standby systems
• Metro Mirror or Global Mirror for data replication
• Automation for workload failover

System Tuning

LPAR Configuration

Adjust LPAR weights:

/T LPAR=LPAR1,WEIGHT=50
/D GRS,RES=(*,*)       /* Display GRS resources */

Workload Manager (WLM)

Service classes:

SERVICE CLASS: CICS_PROD
  IMPORTANCE: 1
  GOALS: RESPONSE(90%,0.2SEC)
  
SERVICE CLASS: BATCH
  IMPORTANCE: 3
  GOALS: VELOCITY(50)

Application environments:

• Define Java, CICS, IMS environments
• Resource limits (CPU, memory, threads)
• Classification rules

System Security

APF (Authorized Program Facility)

APF list (PROGxx):

APF ADD DSNAME(SYS1.USERLIB) VOLUME(SYSR01)
APF DELETE DSNAME(OLD.LOAD.LIB)

Dynamic APF:

SETPROG APF,ADD,DSNAME=SYS1.NEWLIB,VOLUME=VOL001
SETPROG APF,DELETE,DSNAME=OLD.LOAD.LIB

Program Properties Table (PPT)

Define in SCHEDxx:

PPT PGMNAME(MYPROG)
  KEY(7)
  NOSWAP
  PRIV
  PASS

Console Operations

Master Console (MSTR)

Console definition (CONSOLxx):

CONSOLE DEVNUM(001C)
  NAME(MSTCONS)
  AUTH(MASTER)
  CMDSYS(ALL)
  AREA(F)
  ROUTCDE(ALL)
  MSCOPE(*)
  
CONSOLE DEVNUM(001D)
  NAME(SYSTEC)
  AUTH(SYS)

Console Security

RACF console profiles:

RDEFINE OPERCMDS MVS.VARY.TAPE.* UACC(NONE)
PERMIT MVS.VARY.TAPE.* CLASS(OPERCMDS) ID(OPERGRP) ACCESS(UPDATE)
SETROPTS RACLIST(OPERCMDS) REFRESH

System Capacity Planning

Monitoring Points

CPU:

• LPAR utilization
• Service class CPU consumption
• Transaction response times

Storage:

• Real storage usage
• Auxiliary storage usage
• Common area usage

I/O:

• DASD response times
• Channel utilization
• Cache hit ratios

Capacity Modeling

Data collection:

1. RMF data (CPU, storage, I/O) 2. SMF records (workload characteristics) 3. CICS statistics 4. DB2 statistics

Analysis:

• Peak usage times
• Growth trends
• Resource bottlenecks
• Service level compliance

Planning:

• Hardware upgrades
• LPAR weight adjustments
• Workload balancing
• Storage expansion

Troubleshooting

System Dumps

Initiate dump:

/DUMP COMM=('System dump reason')

IPCS (Interactive Problem Control System):

IPCS
SETDEF FILE(dataset) DSNAME('DUMP.DATA')
STATUS
SYSTRACE CT
SUMMARY
VERBEXIT CBFORMAT

Wait States

Common wait codes:

• 001: WTOR (Write To Operator with Reply)
• 021: Master trace buffer full
• 02E: Trace table full
• 091: Unit check on DASD

Performance Issues

High CPU:

/D A,ALL               /* Check active jobs */
/F RMF,D              /* RMF display */
/D PROG,LNKLST        /* Check link list */

Storage shortage:

/D M=STOR             /* Display storage */
/D ASM                /* Display auxiliary storage */
/D GRS,RES=(*,*)      /* Check for serialization */

Best Practices

Daily Operations

1. Review system logs (SYSLOG, OPERLOG) 2. Check job completion status 3. Monitor space utilization 4. Review SMF data for anomalies 5. Verify backup completion

Change Management

1. Test in development environment first 2. Schedule maintenance windows 3. Document all changes 4. Have backout procedures ready 5. Communicate with stakeholders

Security

1. Regular RACF audits 2. Review unauthorized attempts 3. Monitor privileged access 4. Keep security patches current 5. Implement separation of duties

Performance

1. Baseline normal operations 2. Trend analysis 3. Proactive tuning 4. Regular capacity reviews 5. Workload balancing