update config secret name to be more clear

it appears that a change in the cluster-api kubeadm bootstrap provider
is causing secrets with the same name as a machine to be created in the
namespace with the machine object. this is causing the kubeconfig secret
created by the capk controller to be overwritten, which cascades into
kubemark not being able to start.

the fix is to add "kubemark-config" as a suffix to secret.

controllers/kubemarkmachine_controller.go

@@ -61,7 +61,8 @@ import (
 )
 
 const (
-	kubemarkName = "hollow-node"
+	kubemarkName             = "hollow-node"
+	kubemarkSecretNameSuffix = "kubemark-config"
 
 	// MachineControllerName defines the user-agent name used when creating rest clients.
 	MachineControllerName = "kubemarkmachine-controller"
@@ -196,7 +197,7 @@ func (r *KubemarkMachineReconciler) Reconcile(ctx context.Context, req ctrl.Requ
 		}
 		if err := kubemarkClusterClient.Delete(ctx, &corev1.Secret{
 			ObjectMeta: metav1.ObjectMeta{
-				Name:      kubemarkMachine.Name,
+				Name:      secret.Name(kubemarkMachine.Name, kubemarkSecretNameSuffix),
 				Namespace: kubemarkClusterNamespace,
 			},
 		}); err != nil {
@@ -311,7 +312,7 @@ func (r *KubemarkMachineReconciler) Reconcile(ctx context.Context, req ctrl.Requ
 
 	secret := &corev1.Secret{
 		ObjectMeta: metav1.ObjectMeta{
-			Name:      kubemarkMachine.Name,
+			Name:      secret.Name(kubemarkMachine.Name, kubemarkSecretNameSuffix),
 			Namespace: kubemarkClusterNamespace,
 		},
 		Data: map[string][]byte{