I upgraded my GKE clusters to Kubernetes 1.5.6 a couple days ago. I used to be able to scale the l7-default-backend deployment to 3 replicas and increase the CPU resource limits, but now it seems my changes get reverted to their default of 1 replica and 10m CPU limit.

Interestingly enough, when I added the following:

nodeSelector:
    cloud.google.com/gke-nodepool: default-pool

it was persisted no problem.

Here's the current deployment manifest:

apiVersion: extensions/v1beta1
kind: Deployment
metadata:
  annotations:
    deployment.kubernetes.io/revision: "4"
    kubectl.kubernetes.io/last-applied-configuration: '{"kind":"Deployment","apiVersion":"extensions/v1beta1","metadata":{"name":"l7-default-backend","namespace":"kube-system","creationTimestamp":null,"labels":{"k8s-app":"glbc","kubernetes.io/cluster-service":"true","kubernetes.io/name":"GLBC"}},"spec":{"replicas":1,"selector":{"matchLabels":{"k8s-app":"glbc"}},"template":{"metadata":{"creationTimestamp":null,"labels":{"k8s-app":"glbc","name":"glbc"}},"spec":{"containers":[{"name":"default-http-backend","image":"gcr.io/google_containers/defaultbackend:1.0","ports":[{"containerPort":8080}],"resources":{"limits":{"cpu":"10m","memory":"20Mi"},"requests":{"cpu":"10m","memory":"20Mi"}},"livenessProbe":{"httpGet":{"path":"/healthz","port":8080,"scheme":"HTTP"},"initialDelaySeconds":30,"timeoutSeconds":5}}]}},"strategy":{}},"status":{}}'
  creationTimestamp: 2017-03-23T23:30:12Z
  generation: 9
  labels:
    k8s-app: glbc
    kubernetes.io/cluster-service: "true"
    kubernetes.io/name: GLBC
  name: l7-default-backend
  namespace: kube-system
  resourceVersion: "40149922"
  selfLink: /apis/extensions/v1beta1/namespaces/kube-system/deployments/l7-default-backend
  uid: a9772d26-1020-11e7-b9a8-42010af001d0
spec:
  replicas: 1
  selector:
    matchLabels:
      k8s-app: glbc
  strategy:
    rollingUpdate:
      maxSurge: 1
      maxUnavailable: 1
    type: RollingUpdate
  template:
    metadata:
      creationTimestamp: null
      labels:
        k8s-app: glbc
        name: glbc
    spec:
      containers:
      - image: gcr.io/google_containers/defaultbackend:1.0
        imagePullPolicy: IfNotPresent
        livenessProbe:
          failureThreshold: 3
          httpGet:
            path: /healthz
            port: 8080
            scheme: HTTP
          initialDelaySeconds: 30
          periodSeconds: 10
          successThreshold: 1
          timeoutSeconds: 5
        name: default-http-backend
        ports:
        - containerPort: 8080
          protocol: TCP
        resources:
          limits:
            cpu: 10m
            memory: 20Mi
          requests:
            cpu: 10m
            memory: 20Mi
        terminationMessagePath: /dev/termination-log
      dnsPolicy: ClusterFirst
      nodeSelector:
        cloud.google.com/gke-nodepool: default-pool
      restartPolicy: Always
      securityContext: {}
      terminationGracePeriodSeconds: 30
status:
  availableReplicas: 1
  conditions:
  - lastTransitionTime: 2017-04-13T19:19:35Z
    lastUpdateTime: 2017-04-13T19:19:35Z
    message: Deployment has minimum availability.
    reason: MinimumReplicasAvailable
    status: "True"
    type: Available
  observedGeneration: 9
  replicas: 1
  updatedReplicas: 1

How can I scale the l7-default-backend successfully?