More than 10,000 sensors publish data using MQTT to an on-premises Apache Kafka cluster that transforms the data and stores results in Amazon S3. The Kafka server recently crashed and data was lost during recovery. The company needs a highly available, scalable AWS design to avoid future data loss. Which solution satisfies these requirements?

The correct solution leverages AWS IoT Core for highly available and scalable ingestion of MQTT data from 10,000+ sensors. IoT Core natively supports MQTT and can handle massive scale, eliminating the single point of failure and data loss risk associated with the on-premises Kafka cluster. Routing data to Kinesis Data Firehose ensures reliable delivery to Amazon S3, and a Lambda function provides flexible, serverless data transformation. Running EC2 instances for Kafka in active/standby still requires managing the Kafka cluster and doesn't fully address the scalability and operational overhead. Migrating to Amazon MSK is a good option for Kafka, but IoT Core is a more purpose-built and scalable ingestion service for sensor data. Using IoT Core to forward data to a single EC2 instance running Kafka reintroduces a single point of failure and doesn't fully utilize AWS's managed services for high availability and scalability.