Strange behavior adding nodes to seed cluster

Hi!
I currently am trying to evaluate a solution where i use AKS clusters to have a way of running KubeEdge. KubeEdge brings k8s to your IoT device as k8s node. Then you can deploy some workload on those nodes.
What i currently did was:
Create two AKS nodes: main and seed. In the main cluster i installed Kubermatic Kubermetes Platform. I then created a new project and manually added the kubeconfig of the seed cluster to it.
I installed kubeedge on the seed cluster.
My seed cluster has one AKS worker node running the cloudcore of KubeEdge.

$ kubectl get nodes                                 
NAME                                STATUS   ROLES               AGE   VERSION
aks-agentpool-23303655-vmss000000   Ready    agent               41h   v1.19.11
$ kubectl get service                                                                   
NAME        TYPE           CLUSTER-IP    EXTERNAL-IP   PORT(S)                           AGE
cloudcore   LoadBalancer   10.0.41.116   x.x.x.x       10002:30002/TCP,10000:30000/TCP   114m

I’m able to connect a KubeEdge device to the cloudcore and therefore the cluster.
However, all my KubeEdge nodes do have the role master meaning that some master related stuff isrunning on my edge devices. I don’t want this.

$ kubectl get nodes  
NAME                                STATUS   ROLES               AGE   VERSION
aks-agentpool-23303655-vmss000000   Ready    agent               41h   v1.19.11
node0                               Ready    agent,edge,master   60m   v1.19.3-kubeedge-v1.7.1
node1                               Ready    agent,edge,master   27m   v1.19.3-kubeedge-v1.7.1
node2                               Ready    agent,edge,master   27m   v1.19.3-kubeedge-v1.7.1
node3                               Ready    agent,edge,master   27m   v1.19.3-kubeedge-v1.7.1

$ kubectl get pods -o wide                 
NAME                                               READY   STATUS    RESTARTS   AGE    IP            NODE                                NOMINATED NODE   READINESS GATES
coredns-66c464876b-pn6jv-x-kube-system-x-tenant1   0/1     Running   0          64m    172.18.0.2    node0                               <none>           <none>
tenant1-0                                          2/2     Running   0          131m   10.244.0.24   aks-agentpool-23303655-vmss000000   <none>           <none>

Does anyone here can tell me what is going wrong and how to prevent those nodes from becoming masters?

Cheers
Armin