Medium1 markMultiple Choice
Subtask 2.3: Compute SystemsComputeGKEAutoscalingKubernetes

GCP PCA · Question 37 · Compute Systems

You are managing a GKE Standard cluster. During peak hours, the pods are consuming all available CPU on the existing nodes, causing new pods to remain in a 'Pending' state. You need the cluster to automatically add more pods when CPU usage is high, and automatically add more VMs to the cluster when there is no room for new pods. Which TWO autoscaling mechanisms must you configure? (Select TWO)

Answer options:

A.

Horizontal Pod Autoscaler (HPA)

B.

Vertical Pod Autoscaler (VPA)

C.

Cluster Autoscaler

D.

Compute Engine Managed Instance Group Autoscaler

E.

Node Auto-Provisioning (NAP)

How to approach this question

Identify the Kubernetes mechanism for scaling pods, and the mechanism for scaling nodes.

Full Answer

Horizontal Pod Autoscaler (HPA), Cluster Autoscaler
To handle traffic spikes in GKE Standard, you need two layers of autoscaling. The Horizontal Pod Autoscaler (HPA) monitors metrics (like CPU) and increases the number of replica pods. When those new pods cannot fit on the existing VMs, they go into a 'Pending' state. The Cluster Autoscaler detects Pending pods and provisions new Compute Engine VMs (nodes) to accommodate them.

Common mistakes

Selecting the MIG Autoscaler (D). Modifying the underlying MIG of a GKE cluster directly will cause conflicts with Kubernetes.

Practice the full GCP Professional Cloud Architect Practice Exam 6

50 questions · hints · full answers · grading

More questions from this exam