HPA doesn't scale down to minReplicas even though metric is under target for memory and CPU

General Discussions
1

Cluster information

Kubernetes version: 1.12
Cloud being used: (put bare-metal if not on a public cloud) AWS
Installation method: AWS EKS
Host OS: Ubuntu
CNI and version: amazon-k8s-cni:v1.5.3
CRI and version:
metrics-server: v0.3.1

HPA doesn’t scale down to minReplicas even though metric is under target for memory and CPU. It is about 2 hours. The number of replicas is still 7.There is no pending pods. Below is the source code and hpa outputs.

apiVersion: autoscaling/v2beta2
kind: HorizontalPodAutoscaler
metadata:
  name: {{ include "platform.fullname" . }}
  labels:
    app: {{ include "platform.name" . }}
    chart: {{ include "platform.chart" . }}
    release: {{ .Release.Name }}
    heritage: {{ .Release.Service }}
spec:
  maxReplicas: 10
  minReplicas: 1
  metrics:
  - type: Resource
    resource:
      name: cpu
      target:
        averageUtilization: 50
        type: Utilization
  - type: Resource
    resource:
      name: memory
      target:
        averageUtilization: 50
        type: Utilization
  scaleTargetRef:
    apiVersion: extensions/v1beta1
    kind: Deployment
    name: {{ include "platform.fullname" . }}


kubectl get hpa

NAME                         REFERENCE                    TARGETS           MINPODS   MAXPODS   REPLICAS   AGE
platform                 Deployment/platform           44%/50%, 1%/50%         1         10        7       166m


kubectl describe hpa platform

Name:                                                     platform
Namespace:                                                platform-microservices
Labels:                                                   app=platform
                                                          chart=platform-v0.0.1-dev.f7e3ebae0
                                                          heritage=Tiller
                                                          release=platform-microservices
Annotations:                                              <none>
CreationTimestamp:                                        Thu, 12 Dec 2019 10:28:19 -0800
Reference:                                                Deployment/platform
Metrics:                                                  ( current / target )
  resource memory on pods  (as a percentage of request):  44% (473425627428m) / 50%
  resource cpu on pods  (as a percentage of request):     0% (9m) / 50%
Min replicas:                                             1
Max replicas:                                             10
Deployment pods:                                          7 current / 7 desired
Conditions:
  Type            Status  Reason              Message
  ----            ------  ------              -------
  AbleToScale     True    ReadyForNewScale    recommended size matches current size
  ScalingActive   True    ValidMetricFound    the HPA was able to successfully calculate a replica count from memory resource utilization (percentage of request)
  ScalingLimited  False   DesiredWithinRange  the desired count is within the acceptable range
Events:           <none>