fix-crashloop.yaml

# Example Execution Plan: Fix CrashLoopBackOff
# This demonstrates a real-world fix for OOMKilled pods

plan:
  id: example-crashloop-001
  title: "Fix CrashLoopBackOff in api-service (OOMKilled)"
  created: "2026-01-26T13:00:00Z"
  risk: MEDIUM
  estimated_time: "5min"
  
  context:
    issue: "api-service pods in CrashLoopBackOff state"
    root_cause: "Memory limit too low (256Mi), pods getting OOMKilled under load"
    affected_resources:
      - "production/deployment/api-service"
      - "3 pods currently crashing"
  
  rollback:
    method: "Rollback deployment to previous revision"
    commands:
      - "kubectl rollout undo deployment/api-service -n production"
      - "kubectl rollout status deployment/api-service -n production"
    estimated_time: "2min"

steps:
  - id: step-1
    action: kubectl_patch
    description: "Increase memory limit from 256Mi to 512Mi"
    command: |
      kubectl patch deployment api-service -n production -p '{
        "spec": {
          "template": {
            "spec": {
              "containers": [{
                "name": "api",
                "resources": {
                  "limits": {"memory": "512Mi"},
                  "requests": {"memory": "256Mi"}
                }
              }]
            }
          }
        }
      }'
    resource: "production/deployment/api-service"
    risk: MEDIUM
    reversible: true
    expected_outcome: "deployment.apps/api-service patched"
    
  - id: step-2
    action: wait_for_rollout
    description: "Wait for all new pods to be running and ready"
    command: "kubectl rollout status deployment/api-service -n production --timeout=5m"
    resource: "production/deployment/api-service"
    timeout: "5m"
    success_criteria: "deployment 'api-service' successfully rolled out"
    
  - id: step-3
    action: verify
    description: "Verify no pods in CrashLoopBackOff"
    command: "kubectl get pods -n production -l app=api-service -o jsonpath='{.items[*].status.phase}'"
    expected_outcome: "Running Running Running"
    
  - id: step-4
    action: verify
    description: "Check for OOMKilled events in past 5 minutes"
    command: "kubectl get events -n production --field-selector involvedObject.name=api-service --sort-by='.lastTimestamp' | grep OOMKilled | tail -5"
    expected_outcome: "No OOMKilled events after patch"
    
  - id: step-5
    action: verify
    description: "Monitor pod restarts for 30 seconds"
    command: "sleep 30 && kubectl get pods -n production -l app=api-service -o jsonpath='{.items[*].status.containerStatuses[0].restartCount}'"
    expected_outcome: "No new restarts (count unchanged)"

approval:
  required: true
  status: pending
  approver: null
  approved_at: null
  notes: "Awaiting operator approval"

execution:
  started_at: null
  completed_at: null
  duration: null
  status: pending
  steps_completed: 0
  current_step: null
  error: null

validation:
  pre_flight:
    - check: "Deployment api-service exists in production namespace"
      command: "kubectl get deployment api-service -n production"
      passed: null
    - check: "Current memory limit is 256Mi"
      command: "kubectl get deployment api-service -n production -o jsonpath='{.spec.template.spec.containers[0].resources.limits.memory}'"
      passed: null
    - check: "Pods are actually in CrashLoopBackOff"
      command: "kubectl get pods -n production -l app=api-service | grep CrashLoopBackOff"
      passed: null
      
  post_execution:
    - check: "All pods running"
      command: "kubectl get pods -n production -l app=api-service -o jsonpath='{.items[*].status.phase}'"
      expected: "Running Running Running"
      passed: null
    - check: "Memory limit is now 512Mi"
      command: "kubectl get deployment api-service -n production -o jsonpath='{.spec.template.spec.containers[0].resources.limits.memory}'"
      expected: "512Mi"
      passed: null
    - check: "No restarts in last 5 minutes"
      command: "kubectl get pods -n production -l app=api-service --sort-by='.status.startTime' | head -1"
      passed: null