Google Professional Cloud Database Engineer Question 8
Single answerGoogle Cloud PlatformYou are working for an e-commerce company that is planning to migrate its relational database workload to Google Cloud. The current on-premises database handles 50,000 transactions per second (TPS) during peak hours, and you expect a 30% increase in TPS over the next year. The database size is currently 10 TB, and you estimate a 20% annual growth in data storage. The workload is read-heavy, with an 80/20 read-to-write ratio, and low-latency responses are critical. Based on these requirements, which Google Cloud database solution and configuration would you recommend?
- A
Cloud Spanner with sufficient nodes to handle 65,000 TPS and automatic sharding for scalability
- B
Cloud SQL with a high-availability PostgreSQL instance and sufficient storage to handle 12 TB of data
- C
Bigtable with optimized row key design for low-latency reads and pre-provisioned capacity
- D
Firestore in Native Mode with automatic scaling to handle the read-heavy workload
Show answer and explanation
Correct answer: A
Explanation
The workload described in the scenario requires a database solution that supports high TPS, low-latency reads, and scalability for both transactions and storage. Cloud Spanner is the best choice because it is designed to handle these requirements with its horizontal scalability, strong consistency, and relational schema support. It can automatically shard data and scale to meet increased transactional and storage demands, making it ideal for this e-commerce use case.
- A. Correct.
Cloud Spanner is a globally distributed, horizontally scalable, and strongly consistent database that can handle high TPS and low-latency requirements. It also supports automatic sharding, making it suitable for scaling as your workload grows.
- B. Incorrect.
Cloud SQL is a good option for relational database workloads, but it may struggle to scale to the required TPS and low-latency requirements for this use case. It is better suited for smaller-scale workloads.
- C. Incorrect.
Bigtable is a great choice for NoSQL workloads requiring low-latency reads, but it is not designed for relational database use cases such as this one. It does not support SQL querying or relational schemas.
- D. Incorrect.
Firestore in Native Mode is a NoSQL database optimized for hierarchical data and real-time applications, but it is not suitable for handling a relational database workload with the given TPS and data growth requirements.