Google Professional Cloud Database Engineer Question 229
Select 3Google Cloud PlatformYou are designing a highly available and scalable database solution for an e-commerce application hosted on Google Cloud. The application experiences unpredictable traffic spikes, and the database must handle high read and write throughput while minimizing downtime. Which of the following approaches should you take to meet these requirements?
- A
Use Cloud Spanner with multi-region configuration for global availability and horizontal scalability.
- B
Deploy a single PostgreSQL instance on Compute Engine with preemptible VMs to reduce costs.
- C
Set up Cloud SQL with a read replica to handle increased read traffic and failover capability.
- D
Use Firestore in Native mode to leverage automatic scaling and multi-region replication.
- E
Implement a database cluster using self-managed MySQL on Kubernetes Engine with regional failover.
Show answer and explanation
Correct answers: A, C, D
Explanation
To implement a highly scalable and available database solution on Google Cloud, it is crucial to leverage managed services that are designed to handle unpredictable traffic spikes and provide built-in high availability. Cloud Spanner, Cloud SQL with read replicas, and Firestore in Native mode are all managed services that meet these criteria. A single PostgreSQL instance on preemptible VMs lacks high availability, and a self-managed MySQL cluster on Kubernetes requires significant operational effort and does not inherently provide the same level of scalability or availability.
- A. Correct.
Cloud Spanner offers multi-region configurations for high availability and scales horizontally to handle high throughput, making it a suitable choice for global applications with unpredictable traffic.
- B. Incorrect.
A single PostgreSQL instance on preemptible VMs is not highly available as preemptible VMs can be terminated at any time, and there is no redundancy to handle failures.
- C. Correct.
Cloud SQL with a read replica can improve read performance and provides failover capabilities for high availability, making it a good choice for certain workloads.
- D. Correct.
Firestore in Native mode is fully managed, automatically scales based on traffic, and supports multi-region replication for high availability, making it an ideal option for applications with unpredictable traffic.
- E. Incorrect.
While self-managing MySQL on Kubernetes could provide regional failover, it requires significant operational overhead and does not inherently offer the scalability or availability features provided by managed services like Cloud Spanner or Firestore.