You're responsible for an application where numerous IoT devices send data to GCP. These devices are spread globally, and you need to process these messages in real-time for analytics. The devices send data over MQTT. What combination of GCP services should you use?

*Explanation:* To handle data from IoT devices spread globally, particularly data sent over MQTT, while enabling real-time analytics, a suitable combination of GCP services is to use Cloud IoT Core with Pub/Sub for device management and message queuing, coupled with Dataflow for analytics. This approach ensures that you can efficiently manage IoT devices, ingest their data, and process it in real-time for analytics. Other options, such as using Cloud Pub/Sub with MQTT bridge and Cloud Dataflow or App Engine to host an MQTT broker, may not provide the same level of scalability, device management, and real-time analytics capabilities. Cloud Endpoints and Cloud DLP do not directly address MQTT-based IoT data processing, and Cloud Functions triggered by Cloud Storage events are not designed for real-time analytics.