Examkin

Question

Difficulty: HardPlanning and Assessing Cloud Compute Solutions

A digital media publishing company is planning the Google Cloud compute architecture for two newly developed backend workloads:

1. Video Transcoding Batch Pipeline: A fault-tolerant, stateless batch job that processes uploaded media files, checkpoints progress to Cloud Storage, and can handle unexpected server terminations without data loss.
2. User Comments API: A stateless containerized microservice that handles HTTP REST requests for article commentary, experiences unpredictable traffic spikes during peak news events, requires minimal operational management overhead, and must scale down to zero instances when idle to conserve costs.

Which TWO compute architecture strategies should you recommend to meet these functional and operational requirements while minimizing costs?

  1. Deploy the Video Transcoding Batch Pipeline on Compute Engine Spot VMs managed by an instance group.Answer
  2. Deploy the User Comments API microservice onto Cloud Run.Answer
  3. C
    Deploy the User Comments API microservice onto Cloud Functions by packaging the multi-container application into a single event trigger wrapper.
  4. D
    Deploy the Video Transcoding Batch Pipeline on GKE Standard using dedicated, on-demand high-CPU node pools to prevent server interruptions.
  5. E
    Deploy the User Comments API microservice on Google Kubernetes Engine (GKE) Autopilot while configuring custom Linux kernel modules on the cluster nodes.

Answer

Deploy the Video Transcoding Batch Pipeline on Compute Engine Spot VMs managed by an instance group, and deploy the User Comments API microservice onto Cloud Run.
Deploying the video transcoding workload on Compute Engine Spot VMs provides up to 91% cost savings for fault-tolerant batch workloads that store state checkpoints. Deploying the user commentary microservice on Cloud Run fulfills the requirement for containerized deployment, zero infrastructure administration, fast automatic request scaling, and scaling down to zero when idle.

Step-by-Step Solution

1
Analyze the requirements for the Video Transcoding Batch Pipeline.
The pipeline is stateless, fault-tolerant, checkpoints state to Cloud Storage, and can handle interruptions.
Fault-tolerant batch processing workloads that can resume work are prime candidates for Spot VMs, which offer significant cost savings compared to on-demand Compute Engine instances.
2
Analyze the requirements for the User Comments API.
The microservice is containerized, stateless, experiences variable HTTP traffic, requires zero infrastructure overhead, and must scale to zero.
Cloud Run natively supports containerized stateless HTTP applications, automatically scales compute instances based on incoming web requests (including scaling to zero), and eliminates server/node management.
3
Evaluate and eliminate suboptimal compute options.
Cloud Functions is suboptimal for multi-container web microservices; on-demand GKE instances incur unnecessary costs for fault-tolerant jobs; GKE Autopilot restricts custom host kernel modifications.
Choosing the correct serverless and compute abstractions aligns with cost optimization and operational simplicity principles on Google Cloud Platform.

Key Concept

Compute Resource Selection and Trade-Off Assessment (Spot VMs vs Cloud Run vs GKE)
Estimated Time:2m 30s
Rate this question

Topics

Planning and Assessing Cloud Compute Solutions

More questions from this topic