A company wants to migrate an application and an on-premises Apache Kafka cluster to AWS using a replatform approach (not refactor). The application processes incremental updates from an on-premises Oracle database that are published to Kafka. Which managed solution provides the least operational overhead?

Amazon MSK Serverless is the correct choice because it offers the least operational overhead for migrating an existing Apache Kafka cluster. As a serverless solution, AWS automatically manages the capacity, scaling, and patching of the Kafka brokers, eliminating the need for users to provision or manage servers. This aligns perfectly with a replatforming strategy where the goal is to move an application with minimal changes while offloading infrastructure management. Amazon MSK provisioned cluster requires users to manage broker types, scaling, and patching, leading to higher operational overhead. Amazon Kinesis Data Streams is a different streaming service with its own API, requiring application code changes (re-factoring), which goes against the replatforming requirement. Amazon Kinesis Data Firehose is primarily for loading streaming data into data lakes and analytics services, not for hosting a Kafka-compatible message broker.