Question

Difficulty: MediumSelecting and Designing Compute and Application Platforms

A financial media company is migrating two new application components to Google Cloud. Component 1 is a stateless containerized REST API microservice that processes incoming webhooks with highly variable traffic patterns, including extended periods of zero traffic. Component 2 is a containerized nightly data reconciliation task that runs once per day on a schedule and takes approximately 90 minutes to execute. The cloud architecture team requires a design that minimizes operational infrastructure management overhead and avoids paying for idle compute resources. Which TWO compute platform options should you select? (Select TWO.)

  1. Deploy Component 1 to Cloud Run services with autoscaling configured to scale down to zero instances when idle.Answer
  2. Deploy Component 2 as a Cloud Run job triggered on a daily schedule via Cloud Scheduler.Answer
  3. C
    Deploy Component 1 on a dedicated Google Kubernetes Engine (GKE) Standard cluster to guarantee continuous pod readiness.
  4. D
    Deploy Component 2 on a single persistent Compute Engine VM instance that continuously executes a polling loop.

Answer

Deploy Component 1 to Cloud Run services (scaling to zero instances) and deploy Component 2 as a Cloud Run job triggered via Cloud Scheduler.
Cloud Run services provide fully managed hosting for stateless containerized APIs with scale-to-zero capability during zero-traffic windows. Cloud Run jobs efficiently handle containerized tasks that run to completion (up to 24 hours), avoiding idle infrastructure expenses when paired with Cloud Scheduler for invocation.

Step-by-Step Solution

1
Analyze Component 1 requirements (stateless REST API with intermittent traffic).
Cloud Run services provide serverless container hosting that automatically scales down to zero when idle.
This satisfies the requirement for stateless HTTP webhooks while eliminating idle compute charges and infrastructure management.
2
Analyze Component 2 requirements (scheduled batch task running for 90 minutes).
Cloud Run jobs execute containerized batch tasks to completion for up to 24 hours per run.
Combining Cloud Scheduler with Cloud Run jobs allows serverless scheduled batch processing without running dedicated VMs or managing clusters.

Key Concept

Selecting serverless compute platforms (Cloud Run services and Cloud Run jobs) to minimize operational overhead and baseline costs for stateless HTTP workloads and containerized batch tasks.
Rate this question