Chapter 17 Further Reading

IBM Official Documentation

CICS Transaction Server 5.6

  • CICS TS 5.6 Performance Guide (SC34-7385) — The definitive IBM guide to CICS performance tuning. Chapters on dispatcher tuning, storage management, DB2 connection optimization, and workload management are directly relevant to this chapter. The appendix on CICS statistics fields is essential for interpreting SMF 110 records.

  • URL: https://www.ibm.com/docs/en/cics-ts/5.6?topic=performance

  • CICS TS 5.6 System Definition Guide (SC34-7374) — Reference for all SIT parameters discussed in this chapter: MXT, DSALIM, EDSALIM, GCDSALIM, and related storage and task control parameters. The parameter descriptions include valid ranges, defaults, and tuning guidance.

  • CICS TS 5.6 Problem Determination Guide (SC34-7386) — Comprehensive guide to CICS diagnostic tools: CEDF, auxiliary trace, transaction dumps, and system dumps. Chapter 3 covers performance-related problem determination, including the diagnostic methodology referenced in Section 17.6.

  • CICS TS 5.6 CICS Monitoring (SC34-7380) — Detailed description of CICS Monitoring Facility (CMF) data, including SMF 110 record formats, monitoring control tables, and the relationship between CMF data classes (identity, performance, exception). Required reading for anyone building custom performance analysis tools.

  • CICS DB2 Guide (SC34-7373) — The CICS-DB2 connection architecture, including CMDT, THREADLIMIT, THREADWAIT, and the task-related user exit (TRUE) that manages DB2 thread allocation. The section on THREADSAFE programming is essential background for Section 17.2's discussion of L8 TCBs.

  • CICS TS 5.6 Resource Definition Guide (SC34-7375) — Reference for TRANCLASS, TRANSACTION, and PROGRAM resource definitions used throughout this chapter. Includes the CONCURRENCY parameter options (QUASIRENT, THREADSAFE, REQUIRED).

z/OS Integration

  • z/OS MVS Planning: Workload Management (SA38-0689) — The z/OS WLM reference, relevant to Section 17.7's discussion of WLM and CICS interaction. Chapters on service class design, velocity goals, and the enclave model for CICS are directly applicable.

  • z/OS MVS System Management Facilities (SMF) (SA38-0670) — Reference for SMF record types, including the type 110 record structure. Essential for writing custom SMF extraction programs or configuring commercial monitoring tools.

IBM Redbooks

  • CICS Performance Guide (SG24-6530) — The most practical IBM publication on CICS performance tuning. Covers dispatcher analysis, storage tuning, MXT sizing, TRANCLASS design, and DB2 connection optimization with worked examples and real-world benchmarks. Despite its age, the core principles remain applicable to CICS TS 5.6.

  • CICS and DB2: The Best of Both Worlds (SG24-7421) — Integration patterns between CICS and DB2, including thread management, connection pooling, THREADSAFE programming, and the performance implications of different DB2 access patterns from CICS. The CMDT sizing methodology in this Redbook informed the formulas in Section 17.2.

  • CICS Transaction Server from Start to Finish (SG24-8347) — End-to-end CICS guide. Chapter 8 covers performance monitoring and tuning, including practical guidance on statistics collection, CMF configuration, and performance analysis workflows.

  • Threadsafe Considerations for CICS (SG24-6351) — Dedicated Redbook on converting CICS programs to THREADSAFE. Covers prerequisites, testing methodology, common pitfalls, and performance benchmarks. Essential reading before undertaking the THREADSAFE conversion described in Section 17.2 and Case Study 2.

  • CICS and Parallel Sysplex (SG24-6449) — Covers Sysplex-aware CICS features that impact performance: shared temporary storage, coupling facility data tables, named counters, and cross-region workload distribution. Relevant to the capacity planning discussion in Section 17.7.

Technical Articles and Papers

  • "CICS Dispatcher: Under the Hood" — IBM Developer article explaining the QR TCB, open TCB model, and task dispatching in detail. Includes diagrams of the task state machine (running, ready, waiting) and the dispatcher's selection algorithm.

  • "Understanding CICS Storage Domains" — IBM Systems Magazine article providing a visual guide to DSA, EDSA, and GCDSA sub-pools, with practical sizing guidance and SOS prevention strategies.

  • "CICS Performance Analysis Using SMF 110 Records" — SHARE conference presentation covering SMF 110 record extraction, field interpretation, and common performance analysis patterns. Includes sample SAS and REXX code for processing SMF 110 data.

  • "The Art of MAXT Diagnosis" — IBM Support technical document describing the common causes of MAXT conditions, the diagnostic methodology, and the resolution patterns. Includes a decision tree for MAXT triage.

  • "TRANCLASS Best Practices for Production CICS" — IBM CICS blog post covering TRANCLASS design patterns for financial services environments. Includes sample TRANCLASS configurations for ATM, web, and batch-initiated workloads.

Books

  • Horswill, Matthew. CICS: A Practical Guide to System Fine Tuning, 2nd Edition. — The most comprehensive book dedicated to CICS performance tuning. Covers every aspect discussed in this chapter: dispatcher analysis, storage management, MXT sizing, TRANCLASS, and DB2 connection tuning. The worked examples and case studies are drawn from real production environments. Essential reference for any CICS performance specialist.

  • Tull, Sam. CICS Performance and Capacity Tuning. — Focused on capacity planning methodology for CICS environments. Covers workload modeling, growth projection, and the relationship between CICS capacity and z/OS capacity. The capacity planning framework in Section 17.7 draws on principles from this book.

  • IBM. IBM Mainframe Operator's Guide, 9th Edition. — While primarily an operations reference, the chapters on CICS operations (startup, shutdown, monitoring, incident response) provide context for the operational aspects of CICS performance management — particularly the runbook patterns referenced in Section 17.6.

Conference Presentations

  • SHARE Conference — CICS Performance Track — The annual SHARE conference includes dedicated sessions on CICS performance tuning, THREADSAFE conversion, and capacity planning. Presentations from IBM CICS developers often include preview information on upcoming performance features. Archives available at share.org.

  • IBM TechU — CICS Deep Dive Sessions — IBM's technical conference includes multi-hour deep dives on CICS internals, including dispatcher mechanics, storage management, and DB2 connection architecture. The "CICS Performance for Architects" track is particularly relevant to this chapter.

  • GSE (Guide Share Europe) — CICS Working Group — European user group with technical presentations on CICS performance from major European financial institutions. Case studies from banks and insurers with thousands of CICS regions provide real-world validation of the principles in this chapter.

Tools

  • IBM CICS Performance Analyzer (CICS PA) — IBM's primary tool for CICS performance analysis. Processes SMF 110 records and produces reports on transaction performance, resource utilization, and exception conditions. Supports historical trending and cross-region comparison.

  • IBM OMEGAMON for CICS — Real-time CICS monitoring tool that provides graphical dashboards for task counts, storage usage, dispatcher performance, and DB2 connection health. Supports automated alerts based on customizable thresholds.

  • Broadcom CICS Detector (formerly CA CICS Detector) — Commercial CICS performance monitor that provides real-time and historical analysis of CICS transaction performance. Includes automated bottleneck detection and root cause analysis features.

  • BMC MainView for CICS — Commercial monitoring tool with real-time dashboards, automated alerting, and capacity planning features for CICS environments. Supports multi-region correlation for CICSPlex environments.

  • CICS Explorer — IBM's Eclipse-based graphical interface for CICS management. Includes real-time resource monitoring, statistics display, and task management. Free for CICS TS 5.x customers.

  • Chapter 1: z/OS Parallel Sysplex Architecture — Foundation for understanding how CICS regions interact with z/OS dispatching, storage management, and the coupling facility. Required background for the WLM integration discussion in Section 17.7.

  • Chapter 2: z/OS Storage Management — Understanding virtual storage, above/below the bar addressing, and address space layout is prerequisite for the DSA/EDSA/GCDSA discussion in Section 17.3.

  • Chapter 5: z/OS Workload Manager — WLM service classes, velocity goals, and dispatching priority directly affect CICS region performance. The WLM/CICS integration in Section 17.7 builds on Chapter 5's WLM foundations.

  • Chapter 8: DB2 Locking and Concurrency — Lock contention and deadlocks are common causes of CICS task accumulation. The CNB MAXT incident (Section 17.1) was caused by a DB2 drain lock — understanding DB2 locking mechanics (Chapter 8) is essential for diagnosing DB2-induced CICS performance problems.

  • Chapter 11: DB2 Performance Monitoring — DB2 EXPLAIN analysis, DSN_STATEMNT_TABLE, and DB2 statistics are the DB2-side diagnostics that complement the CICS-side diagnostics in Section 17.5. When SMF 110 shows DB2 as the bottleneck, Chapter 11's tools identify the specific DB2 problem.

  • Chapter 13: CICS Architecture — Region topology (TOR/AOR/FOR), MRO, and CICSPlex SM routing are the architectural decisions that determine performance boundaries. This chapter tunes within those boundaries; Chapter 13 sets them.

  • Chapter 15: Channels and Containers — Channel/container performance compared to COMMAREA performance has direct implications for EUDSA consumption and inter-program data passing overhead.

  • Chapter 16: CICS Security — Security checking overhead (RACF calls for transaction security, resource-level security) is a component of per-transaction CPU time. Excessive security granularity can add measurable overhead to high-volume transactions.

  • Chapter 18: CICS Failure and Recovery — Failure detection and recovery mechanisms interact with performance management. Understanding how CICS detects task hangs (DTIMOUT), SOS recovery, and region restart informs the resilience aspects of performance design.

  • Chapter 29: Capacity Planning — Extends the CICS-specific capacity planning in Section 17.7 to the full z/OS environment: MIPS, MSU, storage, DASD, and network capacity across the entire Parallel Sysplex.