Google Professional Cloud Developer Question 4
Select 3Google Cloud PlatformYou are designing an API to handle high-throughput requests for a real-time stock trading application. The API is expected to have low latency and handle spikes in traffic during market hours. Which of the following design decisions will help ensure high performance and scalability?
- A
Implementing a caching layer for frequently requested data using Cloud Memorystore.
- B
Using a global load balancer to distribute incoming requests across multiple regions.
- C
Deploying the API on a single Compute Engine instance with auto-restart enabled.
- D
Asynchronous processing of non-critical tasks using Pub/Sub.
- E
Storing all API response data in a Cloud SQL database for every request.
Show answer and explanation
Correct answers: A, B, D
Explanation
To design a high-performing and scalable API, caching frequently requested data, distributing traffic with a global load balancer, and offloading non-critical tasks to asynchronous processing systems like Pub/Sub are best practices. These approaches reduce latency, increase throughput, and ensure the application can handle traffic spikes. Avoiding practices like relying on a single instance or overloading the database is crucial for maintaining performance and scalability.
- A. Correct.
Implementing a caching layer with Cloud Memorystore reduces the need to repeatedly fetch frequently requested data from the database, improving performance and reducing latency.
- B. Correct.
A global load balancer ensures traffic is distributed across multiple regions, which increases scalability and reduces latency for users in different geographic locations.
- C. Incorrect.
Deploying the API on a single Compute Engine instance creates a single point of failure and limits scalability, even with auto-restart enabled.
- D. Correct.
Asynchronous processing with Pub/Sub offloads non-critical tasks from the API, reducing latency for real-time requests and improving the overall performance of the application.
- E. Incorrect.
Storing all API response data in a Cloud SQL database for every request introduces unnecessary latency and increases database load, which can degrade performance.