Look into the second AutoScaler

: HorizontalPodAutoscaler
apiVersion: autoscaling/v2beta1
metadata:
  name: electroneum
  # namespace: etn2
spec:
  scaleTargetRef:
    kind: Deployment
    name: electroneum
  minReplicas: 15
  maxReplicas: 150
  metrics:
  - type: Object
    object:
      target:
        kind: Service
        name: cluster-metrics
      metricName: cpu_capacity_remaining
      targetValue: 30000