Professional Cloud ArchitectDemo

By Google · Question Mode

QuestionQ53

Analyzing and optimizing technical and business processes

For this question, refer to the KnightMotives Automotive case study. KnightMotives has deployed its new Dealer Tools on Google Kubernetes Engine (GKE), and you are supporting performance of the GKE workloads. Some Pods are stateful.

To prevent these Pods from using an unlimited amount of CPU, you configured resource limits on the Deployments. You determined that the Pods were allocated too much CPU and, at times, consume drastically less. You want to optimize resource use while allowing maximum flexibility if load increases. What should you do?

  • A Configure resource requests per Deployment. Set resource requests slightly above the typical CPU usage observed during monitoring.
  • B Configure resource requests per Deployment. Ensure the requests are equal to the current limits.
  • C Enable vertical Pod autoscaling on each Deployment. Ensure the autoscaling mode is set to Auto.
  • D Enable horizontal Pod autoscaling on each Deployment. Ensure the minimum number of Pods is lower than the currently configured amount in the Deployment.
Explanation

GKE Vertical Pod Autoscaling analyzes workload usage over time and automatically adjusts container CPU and memory resource requests and limits. Its Auto mode applies those adjustments as workload requirements change, which rightsizes over-allocated Pods while retaining capacity to adapt to higher usage. Horizontal Pod Autoscaling changes the number of Pod replicas rather than correcting per-Pod CPU allocation. Google Cloud: Vertical Pod autoscaling

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!