Playbooks
Use these playbooks once you have identified the symptom family. Each playbook is structured around competing hypotheses and evidence collection.
Main Content
flowchart TD
A[Playbooks] --> B[Pod Issues]
A --> C[Connectivity]
A --> D[Node Issues]
A --> E[Operations]
A --> F[Networking]
A --> G[DNS]
A --> H[Network Policy]
A --> I[Scheduling / Placement]
| Category | Documents |
| High-Signal Root Playbooks | Pod CrashLoopBackOff, Node Not Ready, Ingress Not Working, Cluster Autoscaler Issues |
| Pod Issues | Image Pull Failure, CrashLoop, Pending Pods |
| Connectivity | Ingress Failure, Service Unreachable |
| DNS | CoreDNS Query Latency or Drops, External Hostname Resolution Failure |
| Extensions | AGC Traffic Not Flowing, Istio Sidecar Injection Failure, Flux Reconciliation Stuck, Dapr Sidecar Fails to Start |
| Identity | OIDC Issuer Mismatch, Audience Mismatch, RBAC Succeeds but Key Vault Fails, Token Exchange Failure |
| Node Issues | Node Not Ready, CNI IP Exhaustion |
| Network Policy | NetworkPolicy Denies Legitimate Traffic, NetworkPolicy Not Blocking Traffic, Cilium Dataplane Migration Issues |
| Operations | Upgrade Failure, Upgrade Blocked by Pod Disruption Budget, Upgrade Blocked by Deprecated API, Node Image Upgrade Stuck, Surge Upgrade IP Exhaustion, Scaling Failure |
| Scheduling / Placement | Scheduling / Placement, Topology Spread Skew Under Capacity, Ready Capacity Drops Below Desired, Pod Disruption Budget Drain Contract, Availability-Zone-Imbalanced Node Pools and Spread Failures |
| Security | Azure Policy Denies Workload, Defender Alert False Positive, PSS Enforcement Breaks Deployment, Image Signature Verification Failure |
| Scaling | KEDA Scaler Not Triggering, NAP Fails to Provision, HPA Flapping, Spot Eviction Storm |
| Storage | PVC Stuck in Pending, Volume Attach Failure, Volume Mount Failure, Volume Expansion Failure, StatefulSet Stuck During Rolling Update |
| Networking | API Server / kubectl Unreachable, Image Pull Fails in Restricted Egress, Webhook / Control-Plane Calls Blocked, SNAT Port Exhaustion |
See Also
Sources