Test: Module 03 — Async Patterns¶

Instructions: Answer all questions.

Total Points: 37 pts Bonus Available: 5 pts

Section 1: Easy Questions (1 pt each)¶

Q1. What is a compensating transaction and how does it differ from a database rollback?

Q2. Name the two styles of Saga pattern implementation.

Q3. What problem does the Outbox pattern solve?

Q4. In event sourcing, how is current state derived?

Q5. What does CQRS stand for and what does it separate?

Section 2: Medium Questions (2 pts each)¶

Q6. Compare choreography and orchestration sagas. Give one advantage and one disadvantage of each.

Q7. Explain the dual-write problem and why it leads to data inconsistency.

Q8. A saga for a travel booking completes steps 1 (flight reserved) and 2 (hotel booked) successfully, but step 3 (car rental) fails. What must happen next and in what order?

Q9. What are the main trade-offs of event sourcing vs. state-based persistence?

Q10. Why does CQRS usually result in eventual consistency between the read and write models?

Section 3: Hard Questions (3 pts each)¶

Q11. Implement the core of a choreography saga for order fulfillment (inventory → payment → shipping). Show the events published and the compensating logic for each service.

Q12. Implement a Python class OutboxRelay that reads pending outbox rows and publishes them to a mock broker, handling failures idempotently.

Q13. Debug: An orchestration saga always retries step 2 even after it has succeeded. Identify at least 2 possible causes and how to fix each.

Q14. Design the event schema and projection logic for a CQRS read model that shows "orders with total > 00 placed by a given customer in the last 30 days".

Section 4: Expert Questions (5 pts each)¶

Q15. Design a complete saga (orchestration style) for an insurance claim processing workflow with at least 5 steps, including: state machine diagram, compensating transactions, idempotency handling, and timeout/retry strategy.

Q16. Event sourcing and CQRS are often used together. Explain the synergies and the additional complexities this combination introduces, including event schema versioning, projection rebuilding time, and eventual consistency observability.

Bonus¶

Bonus 1. (+5 pts) The "process manager" pattern is sometimes confused with the orchestration saga. Explain the key difference and give an example where a process manager is more appropriate than a saga.