Register an imported cluster using a private registry


I’m in an half-air-gapped environment (I can access the docker registries but only through a proxy) and I have installed the rancher control-plane using our private registry just fine.
Now I’m trying to import an existing cluster and I also need to use our private registry for that.
So i’m creating the cluster in rancher and getting the manifest to apply.
Then, in this manifest I need to modify the image of the containers to use our private registry.
Once done I can apply the manifest using kubectl apply.
This is half working.
As a matter of fact the deployment and the daemon set are created in the first place using the private registry but after that it seems they are updated (by the agent itself?) and reverted to the public image name, so the pods can’t be created.

$ kubectl get all -n cattle-system
NAME                                        READY   STATUS             RESTARTS   AGE
pod/cattle-cluster-agent-657ff69f4f-vkmjk   0/1     ImagePullBackOff   0          157m
pod/cattle-cluster-agent-67ddfc445-7222s    1/1     Running            0          157m 
pod/cattle-node-agent-4j4p8                 1/1     Running            0          157m
pod/cattle-node-agent-ptkzx                 0/1     ImagePullBackOff   0          157m

daemonset.apps/cattle-node-agent   2         2         1       1            1           <none>          157m

NAME                                   READY   UP-TO-DATE   AVAILABLE   AGE
deployment.apps/cattle-cluster-agent   1/1     1            1           157m

NAME                                              DESIRED   CURRENT   READY   AGE
replicaset.apps/cattle-cluster-agent-657ff69f4f   1         1         0       157m
replicaset.apps/cattle-cluster-agent-67ddfc445    1         1         1       157m

If I check the the deployment at this point, it will have the standard image name, and not the one in my private registry:

$ kubectl  -n cattle-system describe deployment.apps/cattle-cluster-agent
Name:                   cattle-cluster-agent
Namespace:              cattle-system
CreationTimestamp:      Tue, 23 Jun 2020 16:47:15 -0400
Labels:                 <none>
Annotations:   2
Selector:               app=cattle-cluster-agent
Replicas:               1 desired | 1 updated | 2 total | 1 available | 1 unavailable
StrategyType:           RollingUpdate
MinReadySeconds:        0
RollingUpdateStrategy:  25% max unavailable, 25% max surge
Pod Template:
  Labels:           app=cattle-cluster-agent
  Service Account:  cattle
      Image:      rancher/rancher-agent:v2.4.3 <= Here I should have
     Port:       <none>
     Host Port:  <none>

Any idea on how I can fix that? Is there any way to configure the ranger-agent image to look for the image at another place?

For sure I can edit again the deployment and deamonsets with the correct image name but I need to do that in a automated way and more straightforward way.
Thanks for you help.