Fix providerID implementation

[Dav-11]

Hello,

I have installed a new cluster using RKE2 and cilium as CNI.

I followed the installation guide and deployed both the controller (https://github.com/apache/cloudstack-kubernetes-provider/blob/main/deployment.yaml) and traefik-ingress-controller.yml.

But the load balancer resource is in pending state:

kubectl -n traefik get services
NAME      TYPE           CLUSTER-IP     EXTERNAL-IP   PORT(S)                      AGE
traefik   LoadBalancer   10.43.184.45   <pending>     80:30836/TCP,443:30387/TCP   8h

The controller logs report this:

E0207 07:44:58.325932       1 controller.go:244] error processing service traefik/traefik (will retry): failed to ensure load balancer: could not find network
I0207 07:44:58.326064       1 event.go:278] Event(v1.ObjectReference{Kind:"Service", Namespace:"traefik", Name:"traefik", UID:"4759be13-4f20-41df-a491-fa51efa1bb10", APIVersion:"v1", ResourceVersion:"44762", FieldPath:""}): type: 'Warning' reason: 'SyncLoadBalancerFailed' Error syncing load balancer: failed to ensure load balancer: could not find network
I0207 07:49:58.326490       1 event.go:278] Event(v1.ObjectReference{Kind:"Service", Namespace:"traefik", Name:"traefik", UID:"4759be13-4f20-41df-a491-fa51efa1bb10", APIVersion:"v1", ResourceVersion:"44762", FieldPath:""}): type: 'Normal' reason: 'EnsuringLoadBalancer' Ensuring load balancer
I0207 07:49:58.326878       1 event.go:278] Event(v1.ObjectReference{Kind:"Service", Namespace:"traefik", Name:"traefik", UID:"4759be13-4f20-41df-a491-fa51efa1bb10", APIVersion:"v1", ResourceVersion:"44762", FieldPath:""}): type: 'Warning' reason: 'UnAvailableLoadBalancer' There are no available nodes for LoadBalancer
E0207 07:49:59.889606       1 controller.go:244] error processing service traefik/traefik (will retry): failed to ensure load balancer: could not find network
I0207 07:49:59.889676       1 event.go:278] Event(v1.ObjectReference{Kind:"Service", Namespace:"traefik", Name:"traefik", UID:"4759be13-4f20-41df-a491-fa51efa1bb10", APIVersion:"v1", ResourceVersion:"44762", FieldPath:""}): type: 'Warning' reason: 'SyncLoadBalancerFailed' Error syncing load balancer: failed to ensure load balancer: could not find network
I0207 07:54:59.890211       1 event.go:278] Event(v1.ObjectReference{Kind:"Service", Namespace:"traefik", Name:"traefik", UID:"4759be13-4f20-41df-a491-fa51efa1bb10", APIVersion:"v1", ResourceVersion:"44762", FieldPath:""}): type: 'Normal' reason: 'EnsuringLoadBalancer' Ensuring load balancer

If I trigger the node labels generation by using

kubectl taint nodes <my-node-without-labels> node.cloudprovider.kubernetes.io/uninitialized=true:NoSchedule

I get errors:

I0207 08:10:24.487510       1 node_controller.go:325] Initializing node mynode0.mydomain.example with cloud provider
E0207 08:10:24.526907       1 node_controller.go:370] failed to initialize node mynode0.mydomain.example at cloudprovider: error fetching node by provider ID: instance not found, and error by node name: instance not found

The cluster is running inside three VMs using an isolated network with the kubernetes network offering.

Can you please help me understanding what went wrong ?