Plan Sync API — Respuesta a Incidentes

# Paso 1: Ver estado de pods
kubectl get pods -n nutriflow

# Paso 2: Si pods en CrashLoopBackOff → revisar logs
# Prereq: kubectl configurado, kubeconfig → cluster prod
# Si falla: aws eks update-kubeconfig --name nutriflow-prod --region eu-west-1
kubectl logs -n nutriflow -l app=plan-sync-api --tail=100

# Paso 3: Revisar historial de deploys
kubectl rollout history deployment/plan-sync-api -n nutriflow

# Paso 4: ROLLBACK si el deploy reciente es sospechoso
kubectl rollout undo deployment/plan-sync-api -n nutriflow

# Paso 5: Escalar réplicas si hay limitación de recursos
kubectl scale deployment/plan-sync-api -n nutriflow --replicas=8

# Paso 6: Verificar recuperación
kubectl rollout status deployment/plan-sync-api -n nutriflow

-- DRY RUN: Contar conexiones idle largas antes de terminar
SELECT count(*)
FROM pg_stat_activity
WHERE state = 'idle'
  AND query_start < now() - interval '10 minutes';

-- Queries lentas activas
SELECT pid, now() - query_start AS duration, query
FROM pg_stat_activity
WHERE state = 'active'
  AND duration > interval '5 seconds'
ORDER BY duration DESC;

-- EJECUTAR solo si count es < 50
SELECT pg_terminate_backend(pid)
FROM pg_stat_activity
WHERE state = 'idle'
  AND query_start < now() - interval '10 minutes';

# Si Redis está saturado, activar circuit breaker
kubectl set env deployment/plan-sync-api \
  REDIS_CIRCUIT_BREAKER_ENABLED=true \
  REDIS_FALLBACK_MODE=passthrough \
  -n nutriflow

# Verificar latencia de dependencias externas
curl -w "Time: %{time_total}s\n" -o /dev/null -s https://api.sendgrid.com/v3/health

# Agrupar errores por tipo
kubectl logs -n nutriflow -l app=plan-sync-api --tail=500 | \
  grep -i error | sort | uniq -c | sort -rn | head -20

# Deshabilitar feature problemática via flag
curl -X POST https://api.nutriflow.io/internal/feature-flags \
  -H "Authorization: Bearer $INTERNAL_TOKEN" \
  -d '{"flag": "SYNC_BULK_MODE", "enabled": false}'

# Revisar cambios recientes en datos

SELECT *
FROM audit_log
WHERE table_name = 'nutrition_plans'
  AND created_at > now() - interval '1 hour'
ORDER BY created_at DESC;

# Ver consumo actual de pods
kubectl top pods -n nutriflow

# Escalar horizontalmente
kubectl scale deployment/plan-sync-api -n nutriflow --replicas=16

# Activar rate limiting
kubectl set env deployment/plan-sync-api \
  RATE_LIMIT_ENABLED=true \
  RATE_LIMIT_RPS=500 \
  -n nutriflow

# Si se sospecha ataque DDoS → aplicar NetworkPolicy
kubectl apply -f network-policy-block-suspicious.yaml