You are previewing IMS 12 Selected Performance Topics.
O'Reilly logo
IMS 12 Selected Performance Topics

Book Description

IBM® Information Management System (IMS™) provides leadership in performance, reliability, and security to help you implement the most strategic and critical enterprise applications. IMS, IMS utilities, and IMS tools continue to evolve to provide value and meet the needs of enterprise customers.

With IMS 12, integration and open access improvements provide flexibility and support business growth requirements. Scalability improvements have been made to the well-known performance, efficiency, availability, and resilience of IMS by using 64-bit storage.

In this IBM Redbooks® publication we provide IMS performance monitoring and tuning information by describing the key IMS performance functions and by showing how to monitor and tune them with traditional and new strategic applications. This book is for database administrators and system programmers.

We summarize methods and tools for monitoring and tuning IMS systems, describe IMS system-wide performance, database, and transaction considerations.

Based on lab measurements, we provide information about recent performance enhancements that are available with IMS 12, and advice about setting performance-related parameters.

Table of Contents

  1. Front cover
  2. Figures
  3. Examples
  4. Tables
  5. Notices
    1. Trademarks
  6. Preface
    1. The team who wrote this book
    2. Now you can become a published author, too!
    3. Comments welcome
    4. Stay connected to IBM Redbooks
  7. Chapter 1. Performance monitoring with IMS utilities
    1. 1.1 IMS utilities
    2. 1.2 IMS and DB Monitor
      1. 1.2.1 Activation and control
      2. 1.2.2 Logging data
    3. 1.3 File Select and Formatting Print utility (DFSERA10)
    4. 1.4 Log Transaction Analysis utility (DFSILTA0)
    5. 1.5 Statistical Analysis utility (DFSISTS0)
    6. 1.6 Fast Path Log Analysis utility (DBFULTA0)
  8. Chapter 2. Performance concepts and monitoring methodology
    1. 2.1 Performance challenges
    2. 2.2 Events for full-function messages
    3. 2.3 Events for Fast Path messages
    4. 2.4 Events for DBCTL
    5. 2.5 Common log records produced for transaction flows
    6. 2.6 Logging in a single IMS system
    7. 2.7 Open transaction flow
    8. 2.8 Monitoring methodology
      1. 2.8.1 Establishing monitoring strategies
      2. 2.8.2 Monitoring multiple systems in DB/DC and DCCTL environments
      3. 2.8.3 Coordinating performance information in an MSC network
      4. 2.8.4 Monitoring Fast Path systems in DB/DC and DCCTL environments
    9. 2.9 Transaction flow in DB/DC and DCCTL environments
      1. 2.9.1 Principal DB/DC and DCCTL monitoring facilities
      2. 2.9.2 Monitoring procedures in a DBCTL environment
    10. 2.10 Monitoring procedures in an open transaction environment
  9. Chapter 3. Performance monitoring tools
    1. 3.1 IMS Performance Analyzer for z/OS, Version 4 Release 3
      1. 3.1.1 Product overview
      2. 3.1.2 Product features
      3. 3.1.3 Forms-based reports
      4. 3.1.4 Message queue transaction and full-function reports
      5. 3.1.5 Fast Path Reports
    2. 3.2 IMS Problem Investigator for z/OS, Version 2 Release 3
      1. 3.2.1 Product overview
      2. 3.2.2 Product features
    3. 3.3 IMS Connect Extensions for z/OS, Version 2 Release 3
      1. 3.3.1 Product overview
      2. 3.3.2 Product features
      3. 3.3.3 Product components
      4. 3.3.4 Product benefits
    4. 3.4 IMS Buffer Pool Analyzer for z/OS, Version 1 Release 3
      1. 3.4.1 Product overview
      2. 3.4.2 Product features
      3. 3.4.3 Reports overview
      4. 3.4.4 Product benefits
    5. 3.5 Tivoli OMEGAMON XE for IMS on z/OS, Version 4 Release 2
      1. 3.5.1 Product overview
      2. 3.5.2 Provided resources
      3. 3.5.3 Product features
    6. 3.6 Transaction Analysis Workbench for z/OS, Version 1 Release 1
      1. 3.6.1 Product overview
      2. 3.6.2 Product features
      3. 3.6.3 Product components
      4. 3.6.4 Product benefits
      5. 3.6.5 Features introduced by Transaction Analysis Workbench
    7. 3.7 Summary
  10. Chapter 4. Platform-related performance functions
    1. 4.1 Dynamic full-function buffer pools
      1. 4.1.1 Performance characteristics of the UPDATE POOL command
      2. 4.1.2 Difference between OSAM and VSAM quiesce processing
    2. 4.2 Increased VSAM pools
    3. 4.3 Logging enhancements
      1. 4.3.1 Striping of OLDS and SLDS
      2. 4.3.2 OLDS buffers specifications
      3. 4.3.3 Migration to buffers above the 2 GB bar
      4. 4.3.4 Migration to striping and buffers in 64-bit storage
    4. 4.4 Disk technology on log performance
      1. 4.4.1 Test environment
      2. 4.4.2 Results
    5. 4.5 WADS redesign
    6. 4.6 System pools in 64-bit real storage
    7. 4.7 64-bit ACB processing
      1. 4.7.1 Defining 64-bit ACB
      2. 4.7.2 Creating and sizing the 64-bit storage pool
      3. 4.7.3 QUERY POOL command
      4. 4.7.4 New 4515 log record
      5. 4.7.5 New monitor records are written in certain situations
    8. 4.8 zAAP or zIIP times in dependent region accounting log records
    9. 4.9 IMS use of zIIP
    10. 4.10 Member online change NAMEONLY option
    11. 4.11 IMS and Workload Manager (WLM)
      1. 4.11.1 WLM terminology
      2. 4.11.2 WLM concepts of service class and classification
      3. 4.11.3 WLM importance
      4. 4.11.4 WLM concepts and goal types
      5. 4.11.5 WLM managed delays
      6. 4.11.6 WLM concepts and performance level
      7. 4.11.7 WLM service class periods
      8. 4.11.8 WLM use for IMS
    12. 4.12 Program load options and IBM Language Environment
      1. 4.12.1 Library routine retention
      2. 4.12.2 Preload library routines
      3. 4.12.3 Preload application programs
      4. 4.12.4 For OS/VS COBOL programs
      5. 4.12.5 Checklist for program-related performance problem
    13. 4.13 Extended address volume for non-VSAM data sets
      1. 4.13.1 Allocating data sets on EAV volumes
      2. 4.13.2 IMS non-VSAM data sets supported
      3. 4.13.3 Example of IMS non-VSAM data sets specification
      4. 4.13.4 Special considerations for HALDB OLR output data sets
    14. 4.14 OTMA support for asynchronous IMS-to-IMS communications
      1. 4.14.1 Usage and benefits of Asynchronous IMS-IMS TCP/IP support
      2. 4.14.2 Performance evaluation results of IMS connect to IMS connect
    15. 4.15 MSC TCP/IP
      1. 4.15.1 MSC TCP/IP uses IMS Connect and the common service layer
      2. 4.15.2 IMS to IMS connect functionality
      3. 4.15.3 IMS considerations
      4. 4.15.4 Setup scenario and switching from VTAM to TCP/IP
      5. 4.15.5 Fallback scenario
      6. 4.15.6 Performance considerations (sample performance test)
      7. 4.15.7 Performance expectation
    16. 4.16 Disabling sysplex serial program management (SSPM)
    17. 4.17 APPC and OTMA shared queue enhancements
      1. 4.17.1 Front-end logging
      2. 4.17.2 IMS DISPLAY commands
      3. 4.17.3 Setup considerations
      4. 4.17.4 APPC and OTMA SQ enhancements benefits
      5. 4.17.5 Evaluation results: APPC synchronous shared queues
      6. 4.17.6 Evaluation results: OTMA synchronous shared queues
    18. 4.18 OTMA ACEE reduction for multiple OTMA clients
      1. 4.18.1 Considerations for OTMA setup
      2. 4.18.2 Benefits
      3. 4.18.3 Evaluation results: OTMA ACEE reduction
    19. 4.19 OTMA performance enhancement
    20. 4.20 IRLM Version 2.3 and 2.2
  11. Chapter 5. Database performance topics
    1. 5.1 Parallel migration to high availability large database (HALDB)
      1. 5.1.1 The problem before the enhancement
      2. 5.1.2 Parallel migration to HALDB
      3. 5.1.3 Parallel migration to HALDB restrictions
    2. 5.2 Archive enhancements
      1. 5.2.1 Log archiving for non-recoverable databases
      2. 5.2.2 New parameter for Log Archive utility (DFSUARC0)
    3. 5.3 Online reorganization performance improvements
      1. 5.3.1 OLR VSAM KSDS sequential access
      2. 5.3.2 Skip GNP call for root-only DB
      3. 5.3.3 Reduce use of the data set busy (ZID) lock during OLR
      4. 5.3.4 Eliminate the block (BID) lock for ILDS updates
      5. 5.3.5 Reduce log records generated during OLR
      6. 5.3.6 OLR locking lookaside
      7. 5.3.7 Increase in concurrent OLR
    4. 5.4 OSAM versus VSAM
      1. 5.4.1 Features and benefits of OSAM for databases
      2. 5.4.2 Considerations for using OSAM
      3. 5.4.3 Summary of benefits of using OSAM
    5. 5.5 Understanding IMS locking for performance
      1. 5.5.1 Lock managers
      2. 5.5.2 Full-function: Database record locks
      3. 5.5.3 Full-function: Segment locks
      4. 5.5.4 Full-function locks: Block locks with IRLM
      5. 5.5.5 Full-function locks: Busy locks
      6. 5.5.6 Fast path locks: CI lock
      7. 5.5.7 Fast path locks: Unit of work (UOW) lock
      8. 5.5.8 Lock timeouts
      9. 5.5.9 Deadlock detection
      10. 5.5.10 Handling deadlock victims
      11. 5.5.11 Design guidance
      12. 5.5.12 Space for lock control blocks
      13. 5.5.13 LOCKMAX usage
      14. 5.5.14 A comparison of PI and IRLM
      15. 5.5.15 Analyze locking with the /TRACE SET program isolation command
      16. 5.5.16 Sample locking reports
    6. 5.6 CICS threadsafe support
    7. 5.7 Temporary close of VSAM data sets with new extents
    8. 5.8 Tuning database buffer pools using Buffer Pool Analyzer
      1. 5.8.1 Function and effect of I/O, storage, and CPU resources
      2. 5.8.2 Buffer pool and subpool organization
      3. 5.8.3 Buffer pool definitions
      4. 5.8.4 Tuning process overview
  12. Chapter 6. Transaction manager performance: Application considerations
    1. 6.1 Full-function transaction
      1. 6.1.1 Transaction response time
      2. 6.1.2 The analysis of response time
      3. 6.1.3 Input time analysis
      4. 6.1.4 Input queue time analysis
      5. 6.1.5 Processing time analysis
      6. 6.1.6 Output queue time analysis
      7. 6.1.7 Monitoring
      8. 6.1.8 Output time analysis
    2. 6.2 Fast Path transaction
      1. 6.2.1 Transaction response time
      2. 6.2.2 The analysis of response time
      3. 6.2.3 Input time analysis
      4. 6.2.4 Input queue time analysis
      5. 6.2.5 Processing time analysis
      6. 6.2.6 Output queue time analysis
      7. 6.2.7 Output time analysis
    3. 6.3 Application considerations
      1. 6.3.1 IMS language interface
      2. 6.3.2 Performance and programming considerations
      3. 6.3.3 IBM Language Environment
  13. Chapter 7. Performance considerations for managing distributed workloads
    1. 7.1 General considerations
    2. 7.2 TCP/IP connectivity through IMS Connect
      1. 7.2.1 Possible performance benefits of IMS Connect
      2. 7.2.2 Routing messages to improve performance
      3. 7.2.3 Recording IMS Connect events
      4. 7.2.4 Real-time capacity management
      5. 7.2.5 Message expiration and its impact on performance
      6. 7.2.6 User ID caching
      7. 7.2.7 Specifying NODELAYACK
    3. 7.3 Open Database
      1. 7.3.1 Challenges of Open Database performance management
      2. 7.3.2 Open Database call-flows and timings
    4. 7.4 Synchronous callout
    5. 7.5 Identifying and managing performance blind spots
  14. Chapter 8. Fast Path performance considerations
    1. 8.1 IMS Fast Path databases
    2. 8.2 DEDB general performance considerations
      1. 8.2.1 I/O IWAIT/CALL
      2. 8.2.2 Elapse time for I/O IWAIT
      3. 8.2.3 Usage of Fast Path resources
    3. 8.3 Lock wait time
    4. 8.4 VSO and SVSO
      1. 8.4.1 Virtual Storage Option (VSO)
      2. 8.4.2 Shared Virtual Storage Option (SVSO)
      3. 8.4.3 Monitoring VSO and SVSO performance
    5. 8.5 Fast Path 64-bit buffer manager
      1. 8.5.1 Fast Path 64-bit buffer manager overview
      2. 8.5.2 Monitoring of Fast Path 64-bit buffer usage
    6. 8.6 Fast Path secondary index (FPSI) and performance comparison with HDAM
    7. 8.7 Fast Path log reduction
  15. Chapter 9. DBRC
    1. 9.1 DBRC background information
    2. 9.2 DBRC performance considerations
      1. 9.2.1 Defining the RECON data sets
      2. 9.2.2 Resolving data set contention problems
    3. 9.3 DBRC RECON maintenance
    4. 9.4 The CLEANUP command parameters
    5. 9.5 DBRC parallel RECON access
    6. 9.6 DBRC execution on Base Primitives Environment (BPE)
  16. Chapter 10. Parallel Sysplex considerations
    1. 10.1 Coupling facility hardware and microcode
      1. 10.1.1 Coupling facility configuration
      2. 10.1.2 Coupling facility microcode
    2. 10.2 Structure sizing
    3. 10.3 IRLM considerations
    4. 10.4 IRLM lock structure
      1. 10.4.1 Lock structure size
      2. 10.4.2 False contention
      3. 10.4.3 Automatic rebuild
      4. 10.4.4 System-managed duplexing
    5. 10.5 VSAM cache structure
    6. 10.6 OSAM cache structure
    7. 10.7 DEDB considerations
    8. 10.8 Application considerations
    9. 10.9 Shared queues
      1. 10.9.1 IMS parameters
      2. 10.9.2 Structure size
      3. 10.9.3 Structure duplexing
      4. 10.9.4 Overflow
      5. 10.9.5 Structure checkpoint
      6. 10.9.6 MVS logger
      7. 10.9.7 FF scheduling differences
      8. 10.9.8 FP Parallel Sysplex processing options
      9. 10.9.9 OTMA and APPC synchronous shared queues enhancements
  17. Chapter 11. Capturing documentation for performance diagnosis
    1. 11.1 Documentation for defect-related software issues
    2. 11.2 Procedures to capture IMS documentation
      1. 11.2.1 The IMS Monitor trace
      2. 11.2.2 IMS in storage table traces
      3. 11.2.3 Running and externalizing IMS lock and DL/I traces
      4. 11.2.4 Storage as a moving target
      5. 11.2.5 The z/OS DUMP command
      6. 11.2.6 GTF trace
      7. 11.2.7 The /DIAGNOSE command
    3. 11.3 Procedure to request a IMS performance health check
  18. Related publications
    1. IBM Redbooks
    2. Other publications
    3. Online resources
    4. Help from IBM
  19. Back cover