Chapter 5 Quiz: z/OS Workload Manager
Test your understanding of WLM concepts, configuration, and diagnostic techniques. Select the best answer for each question.
Question 1. What is the primary difference between WLM goal mode and the older compatibility mode (IPS/IAS)?
A) Goal mode supports more address spaces than compatibility mode B) Goal mode dynamically adjusts dispatching priorities based on performance goals, while compatibility mode uses static priorities C) Goal mode is required for CICS but compatibility mode works for batch D) Goal mode runs only on z/OS, while compatibility mode runs on both MVS and z/OS
Answer
**B.** Goal mode's fundamental innovation is dynamic priority adjustment based on performance goals. Compatibility mode assigned static dispatching priorities through IPS/IAS that could not adapt to changing workload conditions. Goal mode allows WLM to raise or lower priorities based on whether service classes are meeting their goals.Question 2. A service class has a response time goal of 0.30 seconds and the actual average response time is 0.45 seconds. What is the Performance Index (PI)?
A) 0.67 B) 1.00 C) 1.50 D) 0.15
Answer
**C.** PI = Actual / Goal = 0.45 / 0.30 = 1.50. A PI greater than 1.0 means the service class is missing its goal — in this case, performing 50% worse than the target.Question 3. In WLM's importance hierarchy, what happens when an importance-1 service class has a PI of 0.4 (well below its goal) and an importance-2 service class has a PI of 2.0 (well above its goal)?
A) WLM maintains the importance-1 work at maximum priority regardless B) WLM may allocate excess resources from the importance-1 class to help the importance-2 class C) WLM ignores the importance-2 work until all importance-1 goals are exceeded D) WLM raises the importance-2 work to importance 1 temporarily
Answer
**B.** WLM is goal-based, not blindly hierarchical. When importance-1 work is well below its goal (PI 0.4 means performing more than twice as well as needed), WLM will allocate excess resources to help underperforming lower-importance work. WLM does not over-allocate to work that is already meeting its goals.Question 4. Which of the following is NOT a valid WLM goal type?
A) Response time (average) B) Velocity (percentage) C) Discretionary D) Throughput (transactions per second)
Answer
**D.** WLM supports three goal types: response time, velocity, and discretionary. There is no throughput (TPS) goal type. WLM manages resource allocation and priority, not throughput directly. Throughput is an outcome of proper resource management, not a configurable goal.Question 5. A multi-period service class has Period 1 (importance 1, RT 0.25s) and Period 2 (importance 2, velocity 40%). What is the purpose of this configuration?
A) Period 1 applies during daytime, Period 2 applies during the batch window B) Period 1 applies to the first transaction, Period 2 applies to subsequent transactions C) Period 1 applies while a transaction is consuming normal resources; Period 2 kicks in when a transaction consumes more service than expected, demoting it D) Period 1 and Period 2 alternate every 10 seconds during the WLM decision cycle
Answer
**C.** Multi-period service classes manage resource-consuming outliers. A transaction starts in Period 1 at high priority. If it consumes more service units than the period duration threshold, it transitions to Period 2 at a lower importance. This protects the system from runaway transactions without starving them completely.Question 6. At CNB, wire transfer transactions (XFRI, XFRO, XFRW) run in the same CICS region as general inquiry transactions. How does WLM handle the CICS region's dispatching priority?
A) The region runs at the priority of the most recently started transaction B) The region runs at the priority of the highest-priority active task within it C) The region runs at the average priority of all active tasks D) Each task within the region gets an independent dispatching priority from the z/OS dispatcher
Answer
**B.** The CICS region address space runs at the dispatching priority of its highest-priority active task. This means a single wire transfer transaction at importance 1 elevates the entire region, including lower-priority tasks. This is why CNB separates workloads into multiple CICS regions.Question 7. A COBOL CICS transaction uses EXEC CICS LINK to call program CALCBEN. How is CALCBEN classified by WLM?
A) CALCBEN receives its own WLM classification based on the program name B) CALCBEN inherits the calling transaction's WLM service class C) CALCBEN runs at the CICS region's default service class D) CALCBEN is classified based on the LINK command's TRANSID option
Answer
**B.** EXEC CICS LINK executes the target program within the same CICS task. Since WLM classifies at the task level (based on the transaction ID), the linked program inherits the calling transaction's service class. To get a separate classification, you would use EXEC CICS START, which creates a new task.Question 8. What is the primary advantage of WLM-managed initiators over static JES2 initiators?
A) WLM-managed initiators can run more jobs simultaneously B) WLM-managed initiators dynamically allocate batch capacity based on demand and goals C) WLM-managed initiators do not require JCL changes D) WLM-managed initiators are more secure than static initiators
Answer
**B.** WLM-managed initiators start and stop dynamically based on workload demand and whether service classes are meeting their goals. If a critical batch service class is missing its goal, WLM starts additional initiators to increase throughput. If demand drops, initiators drain. This is far more efficient than static allocation.Question 9. What is a WLM enclave?
A) A secure partition within an LPAR for sensitive workloads B) A mechanism that allows work within a subsystem to be independently classified and managed by WLM C) A group of service classes that share a common dispatching priority range D) A Coupling Facility structure used for cross-system WLM coordination
Answer
**B.** An enclave allows work running inside one address space (such as a DB2 stored procedure running inside a WLM-managed SPAS) to be classified to a different service class than the address space itself. This ensures that a batch job calling a stored procedure gets consistent priority treatment across the subsystem boundary.Question 10. Sandra Chen discovered that DB2 stored procedures called by a critical batch job (BATCHCRT, importance 2) were classified at importance 3. The stored procedures accounted for 60% of elapsed time. What was the approximate performance impact?
A) No measurable impact — WLM adjusts automatically B) Approximately 35% increase in elapsed time due to lower priority for the stored procedure portion C) The batch job's importance was effectively reduced to importance 3 for the entire run D) The stored procedures ran at importance 2 because they inherited the calling job's classification
Answer
**B.** As described in the chapter, Sandra found that reclassifying the stored procedures to match the calling job's importance cut elapsed time by 35%. The mismatch meant that 60% of the work was running at lower priority than the batch job itself, creating a bottleneck. The stored procedures do not automatically inherit the calling job's classification — they are classified based on the WLM environment of the DB2 SPAS.Question 11. CNB uses three service policies: DAYTIME, BATCHWIN, and MONTHEND. What changes between DAYTIME and BATCHWIN?
A) Different service classes are defined in each policy B) Different classification rules are used in each policy C) The same service classes are used, but importance levels and goals change D) BATCHWIN disables online processing entirely
Answer
**C.** Service policies change the goals and importance levels of existing service classes without modifying the service classes themselves or the classification rules. BATCHWIN elevates BATCHCRT from importance 3 (daytime) to importance 1, while lowering CICSPROD from importance 1 to importance 2. The classification rules remain identical.Question 12. Which of the following batch job attributes can be used in WLM classification rules?
A) Job name, job class, and accounting information B) Job name, step name, and DD name C) Job name, program name, and PARM value D) Job name, COND code, and return code
Answer
**A.** WLM classification for batch (JES subsystem type) supports job name, job class, accounting information, scheduling environment, user ID, and procedure name. Step name, DD name, program name, PARM values, and return codes are not available as WLM classifiers because WLM classifies at job submission time, before execution details are known.Question 13. An RMF report shows a service class with PI = 0.3. What does this indicate?
A) The service class is performing poorly — only 30% of its goal B) The service class is performing much better than its goal — 3x better C) The service class has used 30% of its allocated resources D) 30% of the transactions in this class are meeting the goal
Answer
**B.** PI = Actual / Goal. For a response time goal, PI = 0.3 means actual response time is 30% of the goal — the service class is performing more than three times better than needed. This indicates potential over-allocation of resources to this class, or that the goal is too lenient.Question 14. Why should you always end each WLM classification group with a wildcard (*) rule?
A) WLM requires a wildcard rule and will reject the definition without one B) Unmatched work defaults to SYSSTC or SYSOTHER at importance 5, which may silently run at minimum priority C) The wildcard rule improves WLM classification performance D) Without a wildcard, unmatched work causes a system ABEND
Answer
**B.** Work that does not match any classification rule defaults to system-defined service classes (SYSSTC for started tasks, SYSOTHER for everything else), which typically run at the lowest importance. A new job or transaction added without updating classification rules will silently run at minimum priority, potentially causing missed SLAs. The wildcard catch-all ensures you control where unclassified work goes.Question 15. In a Parallel Sysplex with four LPARs, how is the WLM service definition managed?
A) Each LPAR has its own independent service definition B) The service definition is shared via the sysplex couple dataset, but dispatching decisions are made per-LPAR C) WLM coordinates dispatching across all LPARs through the Coupling Facility D) The service definition is shared, and dispatching priorities are identical across all LPARs
Answer
**B.** The WLM service definition is stored in the sysplex couple dataset and shared across all systems. Changes propagate automatically. However, dispatching priority *decisions* are made independently on each LPAR based on the local workload mix. A batch-heavy LPAR may assign different priorities than an online-heavy LPAR, even though both use the same service definition.Question 16. What is the approximate dispatching priority range for importance level 1 work?
A) 0–63 B) 64–127 C) 128–191 D) 192–223
Answer
**D.** Importance 1 work occupies the highest WLM-managed priority range, approximately 192–223. Priorities above 223 are reserved for z/OS system components. However, these ranges are approximate and dynamic — WLM adjusts within them based on PI values.Question 17. A DB2 stored procedure is defined with WLM ENVIRONMENT WLMHIGH. What determines the stored procedure's WLM priority?
A) The calling program's service class B) The DB2 main address space's priority C) The service class assigned to the WLMHIGH WLM application environment D) The stored procedure's CPU consumption
Answer
**C.** DB2 stored procedures run in WLM-managed stored procedure address spaces (SPASes). Each SPAS is associated with a WLM application environment (in this case, WLMHIGH), and the service class of that environment determines the priority. The calling program's service class does not automatically transfer to the stored procedure.Question 18. During the WLM decision cycle (approximately every 10 seconds), which of the following actions does WLM take?
A) Terminates low-priority jobs to free resources for high-priority work B) Collects performance data, calculates PIs, and adjusts dispatching priorities C) Recalculates classification rules and reclassifies all active work D) Swaps service policies based on time-of-day settings
Answer
**B.** The WLM decision cycle (policy adjustment interval) collects performance data for all active service classes, calculates Performance Index values, compares them weighted by importance, and adjusts dispatching priorities accordingly. WLM does not terminate work, reclassify active work, or change policies during the decision cycle — those are separate operations.Question 19. Which SMF record type contains WLM workload activity data?
A) SMF Type 30 B) SMF Type 42 C) SMF Type 72 D) SMF Type 89
Answer
**C.** SMF Type 72 records contain WLM performance data. Subtype 3 specifically contains workload activity data (the raw data behind RMF Workload Activity Reports). These records are written at configurable intervals and are essential for historical performance analysis and capacity planning.Question 20. You are diagnosing a slow batch job. The RMF data shows PI = 0.85, dispatching priority = 182, and normal CPU service consumption. What is your diagnosis?
A) WLM is not allocating enough resources — escalate the importance level B) WLM is providing adequate resources — the problem is in the application code, DB2, or I/O, not WLM C) The dispatching priority is too low — check the classification rules D) The service class goal is too aggressive — relax the velocity target