Google ACE Practice Question: A team runs Apache Kafka on self-managed VMs for…

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✓

  Cloud Pub/Sub — Google's managed pub/sub messaging service

  Why this is correct

  Cloud Pub/Sub is the fully managed, natively available messaging service on Google Cloud, providing durable, asynchronous message delivery between independently scaled publishers and subscribers. It supports global topic routing, push/pull subscriptions, at-least-once delivery with configurable ack deadlines, and automatic horizontal scaling to millions of messages per second without any broker infrastructure to patch or monitor. As the direct GCP equivalent to self-managed Kafka, it removes operational overhead while adding native IAM integration and dead-letter topics for failure handling.

• ✗

  Cloud Bigtable — a managed wide-column store for streaming data

  Why it's wrong here

  Cloud Bigtable is a managed, wide-column NoSQL database designed for petabyte-scale analytics and time-series workloads that require extremely low read/write latency. While it can ingest streaming data, it is a storage engine, not a message broker; it cannot buffer and forward events to independent consumers or enforce message ordering and acknowledgment semantics across subscribers. Using Bigtable as a substitute for a pub/sub bus would force you to build polling and distribution logic yourself, making it architecturally wrong for event-driven decoupling.

• ✗

  Cloud Dataflow — the GCP managed streaming processing service

  Why it's wrong here

  Cloud Dataflow is Google Cloud's managed processing engine for both batch and stream data pipelines, built on Apache Beam. It is the correct tool for consuming, transforming, and enriching data from a messaging bus like Pub/Sub, but it does not itself provide the queue or broker functionality; Dataflow acts like Kafka Streams consumer applications, not the Kafka broker that stores and routes messages. Without an underlying messaging service, Dataflow has no input stream to process, which is why it cannot replace the message queue itself.

• ✗

  Cloud Storage — use GCS notification events as a message queue

  Why it's wrong here

  Cloud Storage is a durable, highly available object store for immutable blobs; while it can emit bucket notifications to Pub/Sub on object lifecycle events, those notifications are edge-triggered and not queued messages with retention or subscriber acknowledgments. Treating Cloud Storage as a message queue would require implementing custom polling, lease, and deletion logic to emulate message consumption, and would add significant latency and cost compared to a native pub/sub broker. The notification mechanism is designed to integrate with Pub/Sub, not to be a standalone queue itself.

Go deeper

Related to this question