Check Your Knowledge

Test your understanding of this section’s procedure-autopsy concepts before moving on.

1. Why is a stored procedure a fundamentally different migration problem than a table?
   • A. Procedures are always shorter than table definitions
   • B. A procedure is a stateful process that executes in sequence and can branch, not a single declarative statement
   • C. Procedures never contain any SQL
   • D. Tables require more testing than procedures do
2. Which of the following does NOT have a direct one-to-one equivalent in Spark SQL?
   • A. A SELECT statement with a WHERE clause
   • B. An explicit cursor that fetches and processes rows one at a time
   • C. A JOIN between two tables
   • D. A GROUP BY aggregation
3. In the Autopsy method’s color scheme, what does a yellow block represent?
   • A. Pure set-based SQL
   • B. A transaction boundary (COMMIT/ROLLBACK/SAVEPOINT)
   • C. A session-scoped temp table
   • D. Procedural control flow like IF/ELSE
4. What does a color-coded pass reveal that a plain read of the procedure doesn’t?
   • A. The exact runtime of the procedure in production
   • B. How much of the procedure is genuinely set-based (blue) versus procedural scaffolding
   • C. The original author’s name
   • D. Whether the procedure has ever thrown an error
5. According to “Only the Business Logic Migrates,” which block color is described as almost always business logic?
   • A. Yellow
   • B. Orange
   • C. Blue
   • D. Green
6. Why are green (cursor loop) blocks usually discarded rather than translated literally?
   • A. Because cursors are illegal in Delta Lake
   • B. Because the loop construct itself encodes no business rule – only the statement inside it does
   • C. Because Spark cannot execute any loop-like logic
   • D. Because cursors always contain syntax errors
7. What makes a cascading-trigger package harder to autopsy than a single stored procedure?
   • A. Triggers cannot contain any SQL statements
   • B. The execution order is implicit, scattered across separate trigger definitions, and has to be reconstructed as a graph
   • C. Triggers are always written in a different language than procedures
   • D. Databricks does not allow reading trigger source code
8. In the trigger-graph mapping approach, what does a topological sort of the graph produce?
   • A. A list of every column referenced in any trigger
   • B. The true execution order the cascading triggers implied
   • C. A count of how many triggers exist in the schema
   • D. The original CREATE TRIGGER timestamps
9. On the decomposition worksheet, why is “target pattern” always the last column filled in?
   • A. It’s the least important column
   • B. Committing to a target pattern before classification is finished risks reworking code once a later block doesn’t fit the chosen shape
   • C. Target patterns can only be assigned by a database administrator
   • D. The worksheet template requires columns to be filled alphabetically
10. What is the decomposition worksheet’s role in the migration process?
    • A. A one-time compliance form with no further use
    • B. A per-procedure handoff document connecting the autopsy’s classification work to the actual translation work in the next section
    • C. A replacement for reading the original PL/SQL entirely
    • D. A performance benchmarking report

Answer Key

1. B – A procedure executes statements in sequence, holds state, and branches – properties a single declarative SQL statement doesn’t have.
2. B – Explicit row-at-a-time cursors have no direct equivalent in Spark’s set-based, distributed execution model.
3. B – Yellow marks transaction boundaries like COMMIT, ROLLBACK, and SAVEPOINT.
4. B – The color pass produces a visual map of how much of the procedure is genuinely set-based SQL versus procedural mechanism.
5. C – Blue blocks (pure set-based SQL) are almost always the actual business rules worth preserving.
6. B – The loop mechanism itself is procedural plumbing; only the condition and action inside the loop body carry business meaning.
7. B – Trigger execution order is an emergent property of separate CREATE TRIGGER definitions, discoverable only by reconstructing the dependency graph.
8. B – A topological sort of the trigger dependency graph produces the true execution order the cascade implied.
9. B – Filling in target pattern last ensures classification is complete before any translation commitment is made, avoiding rework.
10. B – The worksheet is the per-procedure artifact that carries the autopsy’s findings forward into the Pattern Translation section’s actual code.

---

| ← Previous: The Decomposition Worksheet | Next: Pattern Translation: Cursors, Triggers, Temp Tables, MERGE → | |:—|—:|