An e-commerce enterprise is architecting a Google Kubernetes Engine (GKE) environment to process core transactional events and inventory updates. The system must meet three strict architectural requirements:
1. High availability: The control plane and worker workloads must withstand a complete single-zone outage without service disruption.
2. Zero node management overhead: The infrastructure team must not manage, patch, or configure underlying Compute Engine virtual machine instances or node pools.
3. Automatic scaling: Compute resources must dynamically scale down when demand drops to eliminate unnecessary costs.
Which GKE cluster architecture configuration best satisfies all these requirements?
- Deploy a Regional GKE Autopilot cluster distributed across multiple availability zones.Answer
- BDeploy a Zonal GKE Standard cluster and configure a single node pool spanning three zones.
- CDeploy a Zonal GKE Standard cluster using Spot VMs across all node pools to automate node lifecycle management and ensure zero downtime.
- DDeploy a Regional GKE Standard cluster and rely on the Horizontal Pod Autoscaler (HPA) to automatically provision and patch worker VM instances.
Answer
Deploying a Regional GKE Autopilot cluster distributed across multiple availability zones satisfies high availability, zero node management overhead, and automated resource scaling.
A Regional GKE Autopilot cluster provides a multi-zonal control plane for fault tolerance against zonal outages, while fully managing node lifecycle, OS patching, and automatic infrastructure scaling to eliminate operational overhead.
Step-by-Step Solution
Key Concept
GKE Autopilot vs. Standard Operational Boundaries and Regional vs. Zonal High Availability Planning