Skip to content

Learning Paths

Use this page to choose a reading path based on your role and goal. Each path is numbered, so read the pages in order for the best result. Every path ends with a checklist of concrete outcomes you should be able to demonstrate.

Pick one primary path first

If you fit multiple roles, pick the one that matches your current goal, complete that path, then read a second path opportunistically. Trying to follow every path in parallel dilutes progress.

Choose Your Path

Role Goal Time Budget Start With
Architect Design cluster topology, networking, identity, and cost model 4-6 hours Overview, Platform Hub
Operator Run AKS in production: cluster ops, upgrades, node pools, monitoring 4-6 hours Prerequisites, Operations Hub
Troubleshooter Diagnose node, pod, ingress, and autoscaler failures 3-5 hours + on-call reference Troubleshooting Hub 
flowchart TD
    A[Start Here] --> B[Platform Concepts]
    B --> C{Choose your path}
    C --> D[Architect: Cluster Design]
    C --> E[Operator: Day-2 Cluster Ops]
    C --> F[Troubleshooter: Fast Diagnosis]
    D --> G[Production Readiness]
    E --> G
    F --> G

Architect Path

Design AKS cluster topology, networking model, identity and secrets flow, and production baseline. Focuses on decisions made once and carried for the life of the cluster.

Time: 4-6 hours

flowchart TD
    A[Overview] --> B[Cluster Architecture]
    B --> C[Networking Models]
    C --> D[Identity and Secrets]
    D --> E[Production Baseline]
    E --> F[Resource Governance]

Read in order:

  1. Overview
  2. AKS vs Other Compute
  3. Platform architecture sequence:
  4. Best Practices sequence:

Outcomes

  • You can pick between Azure CNI, kubenet, and Azure CNI Overlay for a workload profile.
  • You can define a node-pool strategy (system vs user pools, spot vs on-demand) with cost targets.
  • You can design an identity model with Workload Identity, Key Vault CSI, and Azure RBAC.
  • You can produce a production-baseline checklist for a new cluster.

Microsoft Learn anchors

Operator Path

Run AKS in production: cluster creation, node pool operations, version upgrades, credential rotation, and monitoring.

Time: 4-6 hours

flowchart TD
    A[Prerequisites] --> B[Cluster Creation]
    B --> C[Node Pool Operations]
    C --> D[Scaling Operations]
    D --> E[Upgrades and Maintenance]
    E --> F[Monitoring and Rotation]

Read in order:

  1. Prerequisites
  2. Operations sequence:
  3. Best Practices sequence:
  4. Tutorials — hands-on tutorials:

Outcomes

  • You can create a production-grade AKS cluster with a validated node pool topology.
  • You can run a version upgrade with a maintenance window and roll back if needed.
  • You can rotate cluster credentials and validate zero-downtime for critical workloads.
  • You can wire Container Insights and design an alert set for the top failure modes.

Microsoft Learn anchors

Troubleshooter Path

Diagnose fast during a live incident. Focuses on node and pod failure triage, ingress and connectivity issues, and autoscaler behavior.

Time: 3-5 hours + on-call reference

flowchart TD
    A[Architecture Overview] --> B[Decision Tree]
    B --> C[First 10 Minutes]
    C --> D[Playbooks]
    D --> E[Diagnostic Commands]

Read in order:

  1. Troubleshooting Hub
  2. Architecture Overview and Mental Model
  3. Decision Tree and Quick Diagnosis Cards
  4. First 10 Minutes runbooks:
  5. Playbooks Hub — pod CrashLoopBackOff, node not ready, ingress, cluster autoscaler, and more
  6. Reference: Diagnostic Commands
  7. Reference: Limits and Quotas

Outcomes

  • You can run the First 10 Minutes runbook for a node, pod, or connectivity symptom.
  • You can select the right playbook from a symptom description without guessing.
  • You can collect the diagnostic bundle (kubectl, az aks commands) that a playbook expects.
  • You can correlate cluster autoscaler decisions with pod scheduling failures.

Microsoft Learn anchors

Track Selection Matrix

Situation Start with Then continue to
Designing a new AKS platform Architect Path Operator Path
Onboarding to an existing cluster Operator Path Architect Path
Preparing for launch Operator Path Troubleshooter Path
Active incident Troubleshooter Path Operator Path (hardening)

Live incident? Skip the path.

If you are actively responding to a page, jump straight to Troubleshooting Hub, the Decision Tree, and the First 10 Minutes runbooks.

See Also

Sources