Google Professional Cloud Database Engineer Question 18
Single answerGoogle Cloud PlatformYour company is migrating an on-premises transactional database to Google Cloud. The database requires 20,000 IOPS (Input/Output Operations Per Second) for optimal performance and stores 2 TB of data. Which storage and compute configuration should you choose to meet these requirements while being cost-effective?
- A
Use Cloud SQL with a high-memory machine type and standard SSD storage.
- B
Use Cloud SQL with a custom machine type and provisioned IOPS SSD storage.
- C
Use Bare Metal Solution with local SSDs.
- D
Use a managed Cloud Spanner instance with high-compute nodes and regional configuration.
Show answer and explanation
Correct answer: B
Explanation
Cloud SQL with provisioned IOPS SSD storage and a custom machine type is the best choice in this scenario. It allows you to explicitly configure the required performance (20,000 IOPS) while managing costs efficiently. Other options either do not meet the performance requirements or are not cost-effective for a 2 TB transactional database.
- A. Incorrect.
Standard SSD storage cannot provide the required 20,000 IOPS for a high-performance transactional workload. This option would not meet the performance requirements.
- B. Correct.
Provisioned IOPS SSD storage in Cloud SQL allows you to configure the required IOPS explicitly, and using a custom machine type ensures you can optimize the compute resources based on the workload. This is the most cost-effective and performance-compliant solution.
- C. Incorrect.
Bare Metal Solution is designed for specific workloads like Oracle databases, and while it provides high performance, it is not a cost-effective option for this scenario.
- D. Incorrect.
Cloud Spanner is designed for globally distributed databases and is not the best fit for a transactional workload requiring 20,000 IOPS. It also incurs higher costs compared to Cloud SQL for this use case.