KEDA Learning

1. to prevent flapping of replica, HPA by default set stabilization window of 300s if metrics is keep fluctuating. https://kubernetes.io/docs/concepts/workloads/autoscaling/horizontal-pod-autoscale/#stabilization-window
2. to rewrite this behavior, you can configure advanced ScaledObject

   spec:
     scaleTargetRef: {}
     minReplicaCount: 1
     maxReplicaCount: 20
     triggers: []
     advanced:
       horizontalPodAutoscalerConfig:
         behavior:
           scaleDown:
             stabilizationWindowSeconds: 0
             policies:
             - type: Percent
               value: 100
               periodSeconds: 15