Google Professional Cloud Developer Question 71
Select 3Google Cloud PlatformYour team has developed a web application hosted on Google Kubernetes Engine (GKE). You notice that your monthly cloud bill is increasing significantly due to underutilized resources. Which of the following actions could help optimize both cost and resource usage for your GKE cluster?
- A
Enable Cluster Autoscaler to automatically adjust the size of the node pool based on workload demands.
- B
Set resource requests and limits for your pods to ensure efficient utilization of CPU and memory.
- C
Switch from preemptible nodes to standard nodes for your GKE cluster to improve reliability and performance.
- D
Use vertical pod autoscaling (VPA) to dynamically adjust the resource requirements of your pods.
- E
Deploy all applications in a single large pod to reduce the number of nodes required in the cluster.
Show answer and explanation
Correct answers: A, B, D
Explanation
To optimize costs and resource utilization in GKE, you should enable features like Cluster Autoscaler and Vertical Pod Autoscaler, which dynamically adjust resources based on demand. Additionally, properly configuring resource requests and limits avoids over-allocation of resources. Avoid practices that increase costs unnecessarily or reduce scalability, such as using standard nodes without a clear need or deploying all applications in a single pod.
- A. Correct.
Enabling Cluster Autoscaler helps scale the GKE cluster up or down based on workload demands. This ensures that you are not paying for idle or underutilized nodes, optimizing cost and resource usage.
- B. Correct.
Setting resource requests and limits ensures that pods use only the necessary amount of CPU and memory. This avoids over-provisioning and helps better utilize cluster resources, reducing costs.
- C. Incorrect.
Switching from preemptible nodes to standard nodes increases costs since preemptible nodes are cheaper. While it may improve reliability, it does not contribute to cost optimization.
- D. Correct.
Vertical Pod Autoscaling (VPA) optimizes the resource requirements of individual pods by dynamically adjusting their CPU and memory usage, reducing waste and improving resource efficiency.
- E. Incorrect.
Deploying all applications in a single large pod is not a recommended practice as it leads to poor scalability, lack of fault isolation, and resource contention, which can negatively impact performance and cost efficiency.