A data pipeline fails only when two scheduled jobs overlap. Each job succeeds alone, and logs show both consume the same temporary directory.
Question: What should engineers test first?
A Increase storage capacity
B Rewrite both jobs
C Disable all scheduled jobs
D Use separate temporary directories
Answer: D. Use separate temporary directories. Separate directories directly test the shared-resource conflict.
A hospital queue is long even though average service time is low. Arrivals cluster at shift changes, while staff coverage stays flat.
Question: Which change should be tested?
A Stagger staff coverage
B Shorten every appointment
C Add a second waiting room
D Stop measuring arrivals
Answer: A. Stagger staff coverage. Coverage aligned with arrival peaks tests the likely capacity mismatch.
A sensor reports high readings. A reference sensor reads normally, and the high sensor was recently exposed to moisture.
Question: What is the most reliable next step?
A Replace all sensors
B Calibrate after drying
C Average both readings
D Ignore the reference sensor
Answer: B. Calibrate after drying. Drying and calibration test whether moisture caused the suspect reading.
A company reduced an application form’s length, yet completion did not improve. User interviews show people stop when a required document is unavailable.
Question: What should the team address next?
A Shorten the form further
B Change page colors
C Help users obtain documents
D Add optional questions
Answer: C. Help users obtain documents. The unavailable document is the stated blocker, not form length.
A city’s traffic model predicts faster trips, but it assumes drivers will not change routes. A trial showed substantial diversion to side streets.
Question: What should the model include?
A More precise map colors
B Older traffic counts
C Average vehicle size only
D Route-choice feedback
Answer: D. Route-choice feedback. Driver diversion changes the network outcome and must be modeled.
A security patch fixes one vulnerability but breaks an older integration used by a small set of customers. A compatible replacement is available but needs testing.
Question: What is the safest rollout plan?
A Test compatible replacement
B Deploy broken integration
C Skip the security patch
D Remove customer records
Answer: A. Test compatible replacement. Testing the replacement supports security while preserving the required integration.
A service outage recurs after restart. Logs show a memory increase after each request until the system fails.
Question: What should the team investigate?
A Network cable color
B Memory release per request
C User password length
D Restart frequency only
Answer: B. Memory release per request. The growing memory use points to a possible unreleased allocation.
A school’s reading intervention helped some pupils but lowered attendance in another group because sessions conflict with language support.
Question: What should the school change?
A Increase intervention length
B Cancel language support
C Coordinate support schedules
D Apply one schedule universally
Answer: C. Coordinate support schedules. Coordination removes the identified conflict without dropping either service.
A pilot shows a repair method works on new pipes but fails on older materials. The city owns both types and cannot replace all pipes soon.
Question: What is the best deployment rule?
A Apply method everywhere
B Reject all repairs
C Mix both methods randomly
D Use method by pipe type
Answer: D. Use method by pipe type. The evidence supports choosing the repair according to pipe material.
A manufacturing defect appears only in the first hour after each equipment changeover. Operators follow different setup sequences.
Question: What should be standardized first?
A Changeover setup sequence
B Final package design
C Customer delivery route
D Shift meal schedule
Answer: A. Changeover setup sequence. The defect timing points to inconsistent setup during changeover.
A forecast model is accurate overall but consistently underestimates demand during heat waves. Those periods drive shortages.
Question: What should modelers examine?
A Overall average accuracy
B Heat-wave error patterns
C Display font size
D Number of forecast colors
Answer: B. Heat-wave error patterns. Errors during critical heat periods need separate analysis despite good overall accuracy.
A database migration appears successful, but totals differ between old and new systems for a small set of records. The difference may affect billing.
Question: What is the proper next step?
A Delete the old database
B Publish the new totals
C Reconcile affected records
D Assume small errors cancel
Answer: C. Reconcile affected records. Billing-impacting discrepancies must be reconciled before relying on migrated totals.
A city installed a crossing signal, but near-misses continue. Observations show drivers cannot see the signal behind parked delivery vans.
Question: What should be addressed first?
A Add more road markings
B Change pedestrian timing
C Increase the parking fee
D Improve signal visibility
Answer: D. Improve signal visibility. The obstructed signal is the direct cause indicated by observations.
A team’s prototype passed lab tests but fails in the field when temperature fluctuates quickly.
Question: What test should be added?
A Rapid temperature cycling
B More static indoor tests
C Longer user surveys
D Different package labels
Answer: A. Rapid temperature cycling. The missing test should reproduce the changing temperature condition.
A program’s total cost exceeded estimates. The main increase came from repeated emergency purchases when standard stock ran out.
Question: What process should be improved?
A Cut staff training
B Set reorder thresholds
C Remove cost tracking
D Buy only during emergencies
Answer: B. Set reorder thresholds. Reorder thresholds can reduce costly emergency stockouts.
A city’s water-quality issue appears at distant taps but not at the treatment plant. Samples are taken at different times.
Question: What sampling change improves diagnosis?
A Test only the plant
B Use one distant tap
C Sample locations simultaneously
D Stop recording sample times
Answer: C. Sample locations simultaneously. Simultaneous samples help identify where and when quality changes occur.
A team fixes a scheduling bug by adding a delay. The delay hides the error in tests, but users still receive duplicate appointments.
Question: What should the team do?
A Increase the delay
B Remove the test suite
C Rename appointment fields
D Trace duplicate creation logic
Answer: D. Trace duplicate creation logic. The real defect is duplicate creation, so the underlying logic must be traced.
A food program receives enough meals overall, yet one shelter routinely runs short while another has surplus. Distribution happens before current counts arrive.
Question: What process change is most effective?
A Allocate after updated counts
B Increase every shelter’s share
C Stop tracking meal counts
D Send surplus to storage later
Answer: A. Allocate after updated counts. Using updated counts before allocation directs supply where it is needed.
A bridge inspection report lists several defects but does not rank them. One defect affects a support joint and may worsen quickly.
Question: What should engineers do first?
A Repair cosmetic paint
B Assess the support-joint defect
C Close every nearby street
D Delay until annual review
Answer: B. Assess the support-joint defect. A potentially worsening structural defect warrants immediate engineering assessment.
A website accessibility fix passes automated checks but screen-reader users still cannot complete checkout. The issue appears at the payment step.
Question: What should the team do next?
A Repeat automated checks only
B Change homepage images
C Test checkout with users
D Remove payment choices
Answer: C. Test checkout with users. User testing at the failure step can reveal interaction barriers automation missed.
A service redesign improves average completion time but sharply increases failures for users with older devices. The service must work on those devices.
Question: What should be changed?
A Keep average time only
B Remove error reporting
C Require new devices
D Restore older-device support
Answer: D. Restore older-device support. The required compatibility constraint means the redesign must support older devices.
A project’s final result is wrong. Each stage output appears valid alone, but stage order may affect the result.
Question: What is the best debugging method?
A Check interfaces and order
B Rerun only the final stage
C Replace all components
D Ignore intermediate outputs
Answer: A. Check interfaces and order. Checking handoffs and sequence can reveal an integration failure.
A community center’s new booking tool reduced phone calls but produced duplicate reservations when users refreshed the page.
Question: What technical control should be added?
A Add more phone lines
B Make submissions idempotent
C Remove refresh controls
D Extend booking hours
Answer: B. Make submissions idempotent. Idempotent submissions prevent repeated requests from creating duplicate bookings.
A city’s repair team repeatedly fixes the same pothole after rain. The patch material is suitable, but the storm drain beside it is blocked.
Question: What should be repaired first?
A Use more patch material
B Close the entire street
C Clear the storm drain
D Change the road name
Answer: C. Clear the storm drain. The blocked drain allows water to undermine each patch, causing recurrence.
A project team faces several possible fixes and a short deadline. One small test can rule out two causes without changing the live service.
Question: Which approach offers the best information safely?
A Deploy every fix
B Change the live system
C Wait for another failure
D Run the isolated test
Answer: D. Run the isolated test. An isolated test reduces uncertainty without exposing the live service.