An organization is planning a Google Kubernetes Engine (GKE) cluster architecture to host stateless web microservices. The DevOps team wants to eliminate manual node provisioning and operating system patching overhead while ensuring the application remains available even if an entire availability zone fails. Which TWO architectural options should be selected to fulfill these requirements?
- Provision the cluster using GKE Autopilot mode.Answer
- Configure the cluster as a regional cluster.Answer
- CProvision the cluster using GKE Standard mode with default node pools.
- DConfigure Horizontal Pod Autoscaler (HPA) to scale cluster worker nodes when CPU load increases.
- EUse Spot VMs exclusively for a mission-critical non-fault-tolerant database backend to eliminate node management.
Answer
The organization should deploy a regional cluster using GKE Autopilot mode.
Deploying in GKE Autopilot mode shifts node management responsibilities (such as OS updates, node provisioning, and security patches) to Google Cloud. Choosing a regional cluster topology replicates the GKE control plane and worker nodes across multiple availability zones, ensuring continuous availability even if one zone fails.
Step-by-Step Solution
Key Concept
GKE Cluster Operation Modes and Regional High Availability Planning