AWS Certified Data Engineer Associate DEA-C01 Practice Question

Amazon Kinesis Data Streams provides millisecond-level put-to-get latency and, by default, retains records for 24 hours (extendable). Multiple consuming applications can independently read the same shard data, and on-demand mode removes most capacity-planning tasks, keeping operations light.

Amazon Kinesis Data Firehose is designed to deliver data to a destination like Amazon S3, not for real-time consumption by multiple applications. While newer buffering options exist to reduce batching delays, its primary mechanism involves batching for delivery, and it cannot guarantee the required sub-second put-to-get latency. Furthermore, it does not support replaying records. Amazon SQS standard queues do not allow multiple consumers to replay the same messages from a single queue. Running Amazon MSK meets the functional needs but introduces cluster management overhead, which is inconsistent with the requirement to minimize operations.