Question

A logistics company is deploying 200,000 IoT sensors across its delivery fleet. Each sensor emits a small telemetry message (location, temperature, speed) every 5 seconds. The architecture team must design a pipeline that reliably ingests this high-volume, bursty stream, decouples producers from consumers, and feeds both a real-time dashboard and a downstream analytics store. Throughput may grow 10x as the fleet expands, and the solution must scale automatically without manual capacity provisioning. Which Google Cloud design best meets these technical requirements?

Accepted Answer

Publish sensor messages to Pub/Sub, process the stream with Dataflow, and write results to Bigtable for the dashboard and to BigQuery for analytics.

Answer

Have sensors write directly to a Bigtable instance sized for peak load, then run Dataflow batch jobs hourly to populate the dashboard and analytics store.

Answer

Configure sensors to POST to an App Engine endpoint behind an HTTP(S) load balancer, and have the endpoint synchronously write to Cloud SQL.

Answer

Send all messages to a single Compute Engine VM running a message broker, and have consumers poll the VM for new telemetry records.