4569. Optimizer Engineering: plan-stability testing — Build success, failure, concurrency and scale tests. — subscription billing — Go database/sql
DBMS
Published
Updated
0
Views
Verified
Scenario: A subscription billing application must handle plan-stability testing. Build success, failure, concurrency and scale tests. Explain the invariant, SQL strategy, transaction/locking behavior, failure handling, observability and test plan. Use MySQL 8-oriented syntax where applicable.
Looking for the latest
Sarkari Result 2026
updates?
Check the latest government exam
results, merit lists and result updates
on Sarkari Resultess.
Answer & Explanation
4
Answers
Verified
Option 1: Use unparameterized dynamic SQL and rely only on application-side validation for
Use unparameterized dynamic SQL and rely only on application-side validation for plan-stability testing.
Additional Explanation
Correctness depends on explicit invariants, safe SQL, transaction/concurrency handling and verification.
Verified
Option 2: Disable database protections and repair inconsistent rows manually after plan-st
Disable database protections and repair inconsistent rows manually after plan-stability testing runs.
Additional Explanation
Correctness depends on explicit invariants, safe SQL, transaction/concurrency handling and verification.
Verified
Option 3: Run plan-stability testing without defining transaction boundaries, retry behavi
Run plan-stability testing without defining transaction boundaries, retry behavior, or production validation.
Additional Explanation
Correctness depends on explicit invariants, safe SQL, transaction/concurrency handling and verification.
Verified
Correct Answer
Option 4: Use a parameterized, transaction-aware MySQL 8 implementation for plan-stability
Use a parameterized, transaction-aware MySQL 8 implementation for plan-stability testing, validate the invariant, and explicitly test concurrency, retries and recovery.
Additional Explanation
Correctness depends on explicit invariants, safe SQL, transaction/concurrency handling and verification.