Google Professional Cloud Database Engineer Question 6
Single answerGoogle Cloud PlatformYour company is planning to migrate a MySQL database from an on-premises data center to Google Cloud. The current database handles 10,000 queries per second (QPS) during peak usage and has an average storage growth rate of 500 GB per year. You expect a 50% increase in workload within the next two years due to business expansion. The database must ensure high availability and should be able to handle the projected workload with minimal latency. Which Google Cloud database solution and configuration should you choose?
- A
Use Cloud SQL with 1 vCPU and 4 GB RAM, and set up a single regional instance.
- B
Use Cloud SQL with a high-availability configuration and provision 16 vCPUs, 64 GB RAM, and 4 TB of storage.
- C
Use Firestore in Datastore mode with on-demand scaling for automatic workload handling.
- D
Use Cloud Spanner with multi-region configuration and provision nodes based on projected QPS and storage growth.
Show answer and explanation
Correct answer: B
Explanation
Cloud SQL with high-availability configuration is the best choice for this scenario. It can handle the current workload of 10,000 QPS while scaling to accommodate the 50% workload increase over the next two years. The high-availability setup ensures minimal downtime, and provisioning 16 vCPUs, 64 GB RAM, and 4 TB of storage provides sufficient capacity for both current and projected storage needs.
- A. Incorrect.
This option is not sufficient for the current QPS requirements or future workload growth. A single regional instance with 1 vCPU and 4 GB RAM would result in performance degradation under heavy load.
- B. Correct.
This option meets the requirements for high availability, scales well for the expected QPS increase, and provides sufficient resources to handle both current and future workloads efficiently.
- C. Incorrect.
Firestore in Datastore mode is not suitable for relational databases like MySQL. It is designed for NoSQL use cases and cannot handle SQL queries or transactions effectively.
- D. Incorrect.
While Cloud Spanner supports high availability and scalability, it is optimized for globally distributed relational workloads. It may be overkill for this scenario, and its cost might not justify the requirements for the projected workload.