Examkin

Question

Difficulty: EasySelecting and Designing Compute and Application Platforms

An enterprise development team is deploying a stateless REST microservice packaged as a Docker container to Google Cloud. The workload experiences unpredictable traffic spikes followed by extended idle periods. The team requires zero infrastructure management overhead and rapid auto-scaling to zero instances when idle. Which TWO options represent the recommended design choices for this application?

  1. Deploy the containerized service to fully managed Cloud Run to handle request-based scaling and zero-instance idle scale-down.Answer
  2. Configure concurrency settings on Cloud Run so each container instance processes multiple simultaneous requests.Answer
  3. C
    Deploy the microservice onto a dedicated Google Kubernetes Engine (GKE) Standard cluster with a multi-zone node pool.
  4. D
    Deploy the microservice on Compute Engine Managed Instance Groups (MIGs) with scaling policy based strictly on host CPU utilization.
  5. E
    Purchase 3-year Committed Use Discounts (CUDs) for baseline compute capacity to lower costs for the microservice.

Answer

Deploying the microservice to fully managed Cloud Run and configuring container concurrency settings.
Deploying the application to fully managed Cloud Run provides a serverless environment that automatically scales up with HTTP traffic and down to zero instances when idle. Configuring container concurrency allows each instance to handle multiple simultaneous requests efficiently, lowering overall operational costs and management overhead.

Step-by-Step Solution

1
Analyze the workload characteristics and operational constraints.
Identified a stateless, containerized HTTP microservice with bursty, unpredictable traffic, requiring zero-scaling and no infrastructure management.
Serverless platforms like Cloud Run are tailored specifically for stateless web services that need automatic scaling to zero.
2
Select the appropriate serverless compute platform.
Selected Cloud Run because it handles containerized HTTP workloads without requiring server provisioning or GKE cluster maintenance.
Cloud Run abstracts cluster management while automatically scaling based on incoming requests.
3
Configure platform scaling settings for efficiency.
Configured request concurrency per container instance to maximize throughput and resource utilization.
Concurrency allows a single container instance to process multiple requests simultaneously, reducing the total instance count needed during spikes.

Key Concept

Selecting serverless container platforms (Cloud Run) and optimizing scaling configuration for stateless HTTP microservices.
Rate this question

Topics

Selecting and Designing Compute and Application Platforms