Soru

Zorluk: ZorManaging Google Kubernetes Engine Resources

An operations team manages a stateless, fault-tolerant processing service deployed on a Google Kubernetes Engine (GKE) Standard cluster. During peak events, high application traffic causes existing pods to consume maximum allocated CPU resources, while new incoming pods remain in a Pending state due to insufficient cluster compute capacity. The team wants to ensure dynamic scalability at both the application level and infrastructure level while optimizing compute costs. Which TWO management strategies should the team implement? (Select TWO answers.)

  1. Configure a Horizontal Pod Autoscaler (HPA) targeting the Deployment workload to dynamically scale the number of pod replicas based on CPU utilization metrics.Cevap
  2. Enable the GKE Cluster Autoscaler on a node pool configured with Spot VMs to dynamically add node capacity when unschedulable pods are detected.Cevap
  3. C
    Configure a Horizontal Pod Autoscaler (HPA) to automatically provision new Compute Engine worker nodes whenever pod memory request limits are exceeded.
  4. D
    Migrate the cluster to GKE Autopilot mode to enable manual tuning of custom host operating system kernel parameters and sysctl settings on worker nodes.

Cevap

The correct operational strategies are to configure a Horizontal Pod Autoscaler (HPA) to scale pod replicas based on workload metrics and enable the GKE Cluster Autoscaler on a Spot VM node pool to automatically provision worker nodes when pending pods cannot be scheduled.
Deploying a Horizontal Pod Autoscaler (HPA) addresses application load by dynamically increasing or decreasing pod replica counts according to metric thresholds. Enabling GKE Cluster Autoscaler on a node pool consisting of Spot VMs addresses cluster capacity limits by adding infrastructure when unschedulable pods exist, while minimizing compute expenditures for stateless workloads.

Adım Adım Çözüm

1
Analyze workload scaling requirements at the pod replica level.
Configuring a Horizontal Pod Autoscaler (HPA) allows Kubernetes to automatically adjust the target deployment's replica count in response to CPU utilization spikes.
HPA operates within the Kubernetes control plane to scale pod instances up or down based on defined resource thresholds.
2
Analyze infrastructure scaling and cost optimization requirements at the node level.
Enabling GKE Cluster Autoscaler on a Spot VM node pool ensures that new node instances are added automatically when pending pods lack room on existing nodes, while leveraging lower-cost preemptible capacity suited for stateless workloads.
Cluster Autoscaler monitors for unschedulable pods and resizes node pools dynamically, whereas Spot VMs reduce compute costs for fault-tolerant applications.

Anahtar Kavram

Distinguishing between Horizontal Pod Autoscaler (HPA) for scaling pod workload replicas and GKE Cluster Autoscaler for dynamic node pool capacity adjustment.
Bu soruyu puanla