I have a k8s cluster deployed on GKE, there is one "main" nodepool containing 1 node, for all the deployments and one nodepool containing 1 node for kube-ip.

On the main nodepool, I would like to deploy 10 replicas of one of my application (flask API). However GKE is contantly killing my pods when the number exceed 5 to get the pod number back to 5.

I tried to modify the values in my differents yaml files (deployment.yaml and horizontalPodScheduler.yaml)

horizontalPodScheduler.yaml

apiVersion: autoscaling/v2beta1
kind: HorizontalPodAutoscaler
metadata:
  labels:
    app: my-app
    env: production
  name: my-app-hpa
  namespace: production
spec:
  maxReplicas: 20
  metrics:
    - resource:
        name: cpu
        targetAverageUtilization: 80
      type: Resource
  minReplicas: 10
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: my-app

deployment.yaml

apiVersion: apps/v1
kind: Deployment
metadata:
  labels:
    app: my-app
    env: production
  name: my-app
  namespace: production
spec:
  replicas: 10
  selector:
    matchLabels:
      app: my-app
      env: production
  strategy:
    rollingUpdate:
      maxSurge: 25%
      maxUnavailable: 25%
    type: RollingUpdate
  template:
    metadata:
      labels:
        app: my-app
        env: production
    spec:
      nodeSelector:
        cloud.google.com/gke-nodepool: my-pool
      containers:
        - image: gcr.io/my_project_id/github.com/my_org/my-app
          imagePullPolicy: IfNotPresent
          name: my-app-1

Even when I set those values, GKE is always overwritting them to 5 replicas:

Here is the resources summary for my main node, you can see there is plenty of resources to deploy the replicas (it's a pretty simple API)

I also tried to use the "scale" button on GKE UI, but results are the same...