An order-processing system uses choreography: each service reacts to events on a shared EventBridge bus rather than being called directly. The team must capture every event flowing through the bus for later replay and debugging, without adding code to any producer or consumer. Which configuration captures the events for replay?
- ASubscribe an SQS queue to the event bus as a target and let the queue retain the events so the team can re-drive them to consumers during debugging sessions.
- BTurn on detailed CloudWatch metrics for the event bus so every event payload is recorded as a metric and can be replayed from the metric history during debugging.
- CAdd an EventBridge rule that writes each event to a Lambda function which appends payloads to a log file the team can manually re-publish to the bus later.
- DEnable an EventBridge archive on the event bus with an optional pattern, then use the archive to replay archived events back onto the bus when needed. Correct
Why A is wrong: A queue target stores events only until retention expires and re-driving from a queue is not the managed replay EventBridge archive provides for a whole bus.
Why B is wrong: CloudWatch metrics record numeric counts and latencies, not full event payloads, so they cannot store or replay the events that flowed through the bus.
Why C is wrong: A custom Lambda logger adds code and operational burden and lacks the managed replay feature, so it does not meet the no-code archive-and-replay requirement.
Why D is correct: An EventBridge archive stores matching events on the bus and supports replay later, capturing the full stream for debugging without touching producer or consumer code.