This note focuses on design reasoning, failure behavior and operational evidence—the parts that matter in code review, system design and incident response.
Why this problem matters
A consumer may complete business work and fail before committing its offset. Kafka then delivers the record again. That is correct at-least-once behavior. The application must ensure that repeating the message does not create a second charge, shipment or ledger entry.
A useful mental model
Separate message delivery from business effect. Delivery may happen more than once; the durable effect should happen once. Use a stable event ID or business key and make the deduplication decision in the same transaction as the state change whenever possible.
Design principles
The following principles are useful because each one creates a boundary that can be reviewed, tested and observed. They are not independent checkboxes: together they define the behavior of the system under normal load and partial failure.
Assign immutable event IDs at the source
Treat this as an architectural constraint rather than a cleanup item. Put the boundary in code, configuration or the data model so a reviewer can see exactly where it is enforced.
Store processed IDs or enforce a unique business constraint
The benefit becomes visible when timing changes under load or failure. Define the limit explicitly and make the fallback, rejection or recovery behavior observable.
Model state transitions so replaying an already-applied transition is harmless
Ownership matters here. The component that owns the invariant should also own the validation, compatibility rule and operational response when the assumption is violated.
Commit offsets only after durable completion
Prefer the smallest mechanism that preserves correctness. Add sophistication only after measurements show that the simpler design cannot meet the workload.
Choose a deduplication retention period that matches replay expectations
Convert this principle into an automated test, deployment check or runbook step. Otherwise it will drift as dependencies, traffic and team ownership change.
How to validate: Verify the design through duplicate delivery, replay, partition skew, consumer restart and rebalance exercises. Messaging correctness becomes visible only when ownership and delivery are disrupted.
Key trade-offs
Good engineering makes the cost of a choice visible. For this topic, the most important trade-offs are:
| Ordering | Stronger ordering usually reduces available parallelism. |
|---|---|
| Delivery | At-least-once delivery improves durability but requires idempotent effects. |
| Recovery | Retry and replay power must be matched with context, controls and ownership. |
Concrete example
The example below is intentionally small. Its purpose is to expose the control point or data flow that the design depends on, not to present a complete framework implementation.
BEGIN;
INSERT INTO processed_event(event_id) VALUES (?) ON CONFLICT DO NOTHING;
-- if inserted, apply business change
UPDATE payment SET status = 'SETTLED' WHERE id = ? AND status = 'AUTHORIZED';
COMMIT;When applying this pattern, define what happens immediately before and after every durable boundary. That is where duplicate work, stale state, lock duration, timeout overlap or deployment risk usually enters the design.
Common failure modes
Failure modes are more useful than generic “best practices” because they describe the condition the design must survive. Review each one as a concrete test scenario.
- Deduplicating only in memory. The usual consequence is hidden backlog, duplicate work or state that can no longer be explained. Add a bounded guardrail and reproduce the condition under load.
- Using timestamps as unique identifiers. This often passes unit tests because the timing, cardinality or dependency behavior is too clean. Test it with realistic concurrency and an intentionally slow or failing dependency.
- Recording the event as processed before the business transaction commits. During restart or replay, the defect can turn a recoverable incident into inconsistent state. Preserve enough context to detect, stop and safely resume the workflow.
- Assuming “exactly once” at the broker removes downstream database duplication. The safest mitigation is to make the assumption explicit in a constraint, deadline, queue limit or state transition, then alert when the boundary is approached.
What to measure
Production behavior should be visible before a failure becomes a customer complaint. Metrics should connect a technical symptom to a workload, business state or recovery objective.
- Duplicate event rateUse this as an early saturation signal and define what healthy, warning and overloaded behavior look like.
- Deduplication-table growthBreak this down by service version, endpoint, partition or tenant so aggregate averages do not hide one failing path.
- State-transition no-op countCorrelate this with user-visible latency and error rate to distinguish harmless internal work from customer impact.
- Offset commit failuresTrack both the level and the age of the condition; an old small backlog can be more serious than a brief large spike.
- Replay success during recovery exercisesReview this after deployments and failure drills so the dashboard proves recovery, not only steady-state health.
Interview-ready explanation
A strong explanation starts with the invariant: state what must remain true even when requests repeat, dependencies slow down or instances restart. Then describe the mechanism that preserves it, the failure mode that mechanism introduces and the signal that proves it is working.
For At-Least-Once Delivery and Idempotent Consumers, avoid listing tools first. Explain the workload and boundary, walk through the normal path, introduce one realistic failure and show how the system recovers. Finish with the metric or test that validates the claim. That structure demonstrates senior engineering judgment more clearly than naming patterns without context.
Review checklist
Use this checklist during design review, implementation planning or incident follow-up:
- Use stable IDs.
- Make dedup and mutation atomic.
- Define replay behavior.
- Retain dedup state appropriately.
- Test crash points before and after commit.
A sound design is not the one with the most patterns. It is the one whose invariants, limits and recovery paths are explicit—and can be demonstrated.