Platform¶
This section explains how AKS works so you can design clusters with realistic assumptions about networking, identity, storage, ingress, and scale.
Main Content¶
graph TD
A[Cluster Architecture] --> B[Node Pools]
B --> C[Networking Models]
C --> D[Ingress and Load Balancing]
D --> E[Identity and Secrets]
E --> F[Storage Options]
F --> G[Scaling]| Document | Description |
|---|---|
| Cluster Architecture | Control plane, worker nodes, system components, and Azure resource relationships |
| Node Pools | System vs user pools, OS choices, scheduling, and lifecycle boundaries |
| Networking Models | Azure CNI Overlay, Azure CNI Pod Subnet, and Kubenet trade-offs |
| Ingress and Load Balancing | Ingress controllers, Services, public/private exposure, and edge paths |
| Identity and Secrets | Microsoft Entra ID, workload identity, managed identity, and Key Vault integration |
| Storage Options | Persistent volumes, CSI drivers, Azure Disks, Azure Files, and secret mounts |
| Scaling | Pod and node autoscaling, workload sizing, and cluster growth controls |
Advanced Topics¶
- Separate platform decisions into cluster baseline, workload baseline, and environment baseline.
- Keep a cluster architecture decision record for network plugin, ingress pattern, and secret strategy.