What is the difference between at-least-once and exactly-once delivery?
Answer
Distributed systems coordinate work across independent machines where messages can be delayed, duplicated, or lost. Design for partial failure with timeouts, idempotency, retries, and explicit consistency guarantees.
Example
Consider a production distributed systems change: define the expected behavior and failure modes first, implement the smallest observable change, then verify it with a focused test or measurement. The right implementation depends on the system boundary and its constraints.
Practical considerations
- Choose the approach from the requirement and constraints, not from habit.
- Include validation, error handling, and cleanup where the boundary requires them.
- Verify the observable result with focused tests or measurement.
Follow-up prompts
- What failure mode would you expect if this were implemented incorrectly?
- How would you test this behaviour?
- What changes when the feature must scale to a larger application or team?
In practice
For this easy-level topic, make assumptions explicit, choose the smallest safe implementation, and verify the behavior at the relevant boundary.