Skip to content

Troubleshooting

This section is a hypothesis-driven AKS troubleshooting hub. Use it to move from symptom to evidence, then to a focused playbook.

Main Content

flowchart TD
    A[Observe Symptom] --> B[Decision Tree]
    B --> C[First 10 Minutes]
    C --> D[Playbook]
    D --> E[Diagnostic Commands]
    E --> F[Validated Root Cause]
Need Start Here
Understand AKS failure surfaces Architecture Overview
Route a symptom quickly Decision Tree
Know what evidence to gather Evidence Map
Apply a mental framework Mental Model
Jump to symptom cards Quick Diagnosis Cards
Respond in the first minutes First 10 Minutes
Execute detailed runbooks Playbooks
Run reusable log investigations KQL Query Packs

Quick Routing Areas

  • Pods: image pulls, crashes, scheduling failures.
  • Connectivity: ingress routing, Services, DNS, and egress.
  • Nodes: readiness, pressure, and IP exhaustion.
  • Operations: upgrade failures, scaling failures, and maintenance side effects.
  • Control plane: API server latency, audit denies, autoscaler decisions, and webhook issues.
  • Storage: PVC binding, attach failures, and CSI-controller evidence.

See Also

Sources