Skip to content

Portal | Level: L1: Foundations | Topics: Kubernetes Core | Domain: Kubernetes

kubectl Debugging Cheat Sheet

Dense command reference grouped by scenario. One-line explanations. Cleanup snippets included.

For deeper coverage: training/interactive/exercises/ (k8s track)

Pod Status & Lifecycle

kubectl get pods -n grokdevops                              # List pods with status
kubectl get pods -n grokdevops -o wide                      # Include node + IP
kubectl get pods -n grokdevops -w                           # Watch for changes
kubectl get pods -A                                         # All namespaces
kubectl get pods -n grokdevops --sort-by=.status.startTime  # Sort by start time
kubectl get pods -n grokdevops -l app.kubernetes.io/name=grokdevops  # By label

Pod Logs

kubectl logs -n grokdevops deploy/grokdevops                # Current logs
kubectl logs -n grokdevops deploy/grokdevops --previous     # Previous container (after crash)
kubectl logs -n grokdevops deploy/grokdevops -f             # Follow/stream
kubectl logs -n grokdevops deploy/grokdevops --tail=100     # Last 100 lines
kubectl logs -n grokdevops deploy/grokdevops --since=5m     # Last 5 minutes
kubectl logs -n grokdevops -l app.kubernetes.io/name=grokdevops --all-containers  # All pods

Describe & Events

kubectl describe pod -n grokdevops -l app.kubernetes.io/name=grokdevops    # Full pod detail
kubectl describe deployment grokdevops -n grokdevops                        # Deployment detail
kubectl get events -n grokdevops --sort-by='.lastTimestamp'                 # Recent events
kubectl get events -n grokdevops --field-selector reason=Failed             # Failed events only
kubectl get events -n grokdevops --field-selector reason=OOMKilling         # OOM events

Exec Into Pods

kubectl exec -n grokdevops deploy/grokdevops -- sh                         # Shell into pod
kubectl exec -n grokdevops deploy/grokdevops -- cat /etc/resolv.conf       # Check DNS config
kubectl exec -n grokdevops deploy/grokdevops -- wget -qO- http://localhost:8000/health  # Test endpoint
kubectl exec -n grokdevops deploy/grokdevops -- env                        # Check env vars
kubectl exec -n grokdevops deploy/grokdevops -- ps aux                     # Check processes

Debug Pods (Ephemeral)

# Run a debug pod in the namespace
kubectl run debug -n grokdevops --rm -i --restart=Never --image=busybox:1.36 -- sh

# DNS test
kubectl run dns-test -n grokdevops --rm -i --restart=Never --image=busybox:1.36 -- nslookup grokdevops

# Curl test
kubectl run curl-test -n grokdevops --rm -i --restart=Never --image=curlimages/curl -- curl -s http://grokdevops/health

Deployments & Rollouts

kubectl get deploy -n grokdevops                                          # List deployments
kubectl rollout status deployment/grokdevops -n grokdevops                # Rollout status
kubectl rollout history deployment/grokdevops -n grokdevops               # Revision history
kubectl rollout undo deployment/grokdevops -n grokdevops                  # Rollback to previous
kubectl rollout restart deployment/grokdevops -n grokdevops               # Restart all pods
kubectl scale deployment grokdevops -n grokdevops --replicas=3            # Scale manually
# Undo scale: kubectl scale deployment grokdevops -n grokdevops --replicas=1

Services & Endpoints

kubectl get svc -n grokdevops                                             # List services
kubectl get endpoints -n grokdevops                                       # Check endpoints (IPs)
kubectl describe svc grokdevops -n grokdevops                             # Service detail
kubectl port-forward -n grokdevops svc/grokdevops 8000:80                 # Port forward
# Stop: Ctrl+C

Resource Usage (requires metrics-server)

kubectl top nodes                                                         # Node CPU/memory
kubectl top pods -n grokdevops                                            # Pod CPU/memory
kubectl top pods -n grokdevops --sort-by=memory                           # Sort by memory
kubectl top pods -A                                                       # All namespaces

HPA (Horizontal Pod Autoscaler)

kubectl get hpa -n grokdevops                                             # HPA status
kubectl describe hpa grokdevops -n grokdevops                             # HPA detail + events
kubectl get hpa -n grokdevops -w                                          # Watch scaling

ConfigMaps & Secrets

kubectl get configmap -n grokdevops                                       # List ConfigMaps
kubectl describe configmap <name> -n grokdevops                           # View ConfigMap
kubectl get secret -n grokdevops                                          # List secrets
kubectl get secret <name> -n grokdevops -o jsonpath='{.data.key}' | base64 -d  # Decode secret

Helm

helm list -n grokdevops                                                   # List releases
helm history grokdevops -n grokdevops                                     # Release history
helm status grokdevops -n grokdevops                                      # Release status
helm get values grokdevops -n grokdevops                                  # Current values
helm get manifest grokdevops -n grokdevops                                # Rendered manifests
helm template grokdevops devops/helm/grokdevops -f devops/helm/values-dev.yaml --debug  # Debug template
helm rollback grokdevops 0 -n grokdevops                                  # Rollback to previous

Observability Stack

# Prometheus
kubectl port-forward -n monitoring svc/kube-prometheus-stack-prometheus 9090:9090
# → http://localhost:9090/targets

# Grafana
kubectl port-forward -n monitoring svc/kube-prometheus-stack-grafana 3000:80
# → http://localhost:3000 (admin/prom-operator)

# Loki
kubectl port-forward -n monitoring svc/loki 3100:3100
curl http://localhost:3100/ready

# Check all monitoring pods
kubectl get pods -n monitoring

# ServiceMonitor
kubectl get servicemonitor -n grokdevops -o yaml

Network Debugging

kubectl get networkpolicy -n grokdevops                                   # List policies
kubectl describe networkpolicy -n grokdevops                              # Policy detail
kubectl get ingress -n grokdevops                                         # List ingress rules
kubectl describe ingress -n grokdevops                                    # Ingress detail

# Test connectivity from pod
kubectl exec -n grokdevops deploy/grokdevops -- wget -qO- --timeout=3 http://grokdevops/health

RBAC

kubectl auth can-i --list -n grokdevops                                   # What can I do?
kubectl auth can-i get pods -n grokdevops                                 # Specific check
kubectl auth can-i get pods --as=system:serviceaccount:grokdevops:default -n grokdevops  # As SA
kubectl get roles,rolebindings -n grokdevops                              # Namespace RBAC
kubectl get clusterroles,clusterrolebindings | grep grokdevops            # Cluster RBAC

JSON Path & Output Formatting

kubectl get deploy grokdevops -n grokdevops -o yaml                       # Full YAML
kubectl get deploy grokdevops -n grokdevops -o json                       # Full JSON
kubectl get pods -n grokdevops -o jsonpath='{.items[*].metadata.name}'    # Pod names
kubectl get deploy grokdevops -n grokdevops -o jsonpath='{.spec.template.spec.containers[0].image}'  # Image
kubectl get pods -n grokdevops -o jsonpath='{range .items[*]}{.metadata.name}{"\t"}{.status.phase}{"\n"}{end}'  # Name + phase table
kubectl get pods -n grokdevops -o custom-columns='NAME:.metadata.name,STATUS:.status.phase,RESTARTS:.status.containerStatuses[0].restartCount'  # Custom columns

jq & yq Patterns

# jq: filter and transform JSON output
kubectl get pods -n grokdevops -o json | jq -r '.items[].metadata.name'                    # Pod names
kubectl get pods -n grokdevops -o json | jq '.items[] | select(.status.phase != "Running")'  # Non-running pods
kubectl get pods -n grokdevops -o json | jq -r '.items[] | "\(.metadata.name)\t\(.status.phase)"'  # Name + phase
kubectl get pods -n grokdevops -o json | jq '[.items[] | .status.containerStatuses[]? | select(.restartCount > 0)] | length'  # Count pods with restarts
kubectl get events -n grokdevops -o json | jq -r '.items | sort_by(.lastTimestamp) | .[-5:] | .[].message'  # Last 5 event messages
kubectl get deploy -n grokdevops -o json | jq -r '.items[] | "\(.metadata.name)\t\(.spec.replicas)\t\(.status.readyReplicas // 0)"'  # Deploy replica status

# yq: filter and transform YAML output directly (no JSON round-trip)
kubectl get deploy grokdevops -n grokdevops -o yaml | yq '.spec.template.spec.containers[0].image'  # Image
kubectl get deploy grokdevops -n grokdevops -o yaml | yq '.spec.template.spec.containers[0].resources'  # Resource limits
kubectl get cm -n grokdevops -o yaml | yq '.items[].metadata.name'                         # ConfigMap names
kubectl get svc grokdevops -n grokdevops -o yaml | yq '.spec.ports'                        # Service ports
kubectl get deploy grokdevops -n grokdevops -o yaml | yq '.spec.template.spec.containers[0].env'  # Env vars

# yq: edit manifests in-place (useful for patching files before apply)
yq -i '.spec.replicas = 3' deployment.yaml                                                 # Change replica count
yq -i '.metadata.labels.version = "v2"' deployment.yaml                                    # Add/update label
yq -i 'del(.spec.template.spec.containers[0].resources.limits)' deployment.yaml            # Remove limits

# Combine: diff two deployments
diff <(kubectl get deploy app-v1 -n grokdevops -o yaml | yq 'del(.metadata.managedFields)') \
     <(kubectl get deploy app-v2 -n grokdevops -o yaml | yq 'del(.metadata.managedFields)')

Cleanup & Nuclear Options

# Redeploy app from Helm (resets all patches)
helm upgrade grokdevops devops/helm/grokdevops -n grokdevops -f devops/helm/values-dev.yaml

# Remove everything this repo deployed
make undeploy-all

# Delete stuck pods
kubectl delete pod <name> -n grokdevops --grace-period=0 --force

# Remove chaos artifacts
kubectl delete networkpolicy -n grokdevops -l chaos=true
kubectl delete pod -n grokdevops -l chaos

Wiki Navigation

Related Content