ctx,
		mock.IsType(pod),
		util.GetPodIPsIfPresent(pod),
		"",
	).Return(nil)

	server := getFakeDP(fs, client.Kube())

	handlers := setupHandlers(ctx, client, server, "istio-system")
	client.RunAndWait(ctx.Done())
	go handlers.Start()

	// label the namespace
	labelsPatch := []byte(fmt.Sprintf(`{"metadata":{"labels":{"%s":"%s"}}}`,
		label.IoIstioDataplaneMode.Name, constants.DataplaneModeAmbient))

	_, err := client.Kube().CoreV1().Namespaces().Patch(ctx, ns.Name,
		types.MergePatchType, labelsPatch, metav1.PatchOptions{})
	assert.NoError(t, err)

	client.RunAndWait(ctx.Done())

	payload, _ := json.Marshal(valid)

	// serialize our fake plugin event
	addEvent, err := processAddEvent(payload)
	assert.Equal(t, err, nil)

	// Push it thru the handler

	}

	rc, err := NewRepairController(client, cfg)
	if err != nil {
		repairLog.Fatalf("Fatal error constructing repair controller: %+v", err)
	}
	go rc.Run(ctx.Done())
	client.RunAndWait(ctx.Done())
}

// Set up Kubernetes client using kubeconfig (or in-cluster config if no file provided)
func clientSetup() (kube.Client, error) {
	config, err := kube.DefaultRestConfig("", "")
	if err != nil {

Running the controller actually starts processing things.
Normally, this just means running the queue.
All informers created by `kube.Client` are kept track in the client, and started in one go with `RunAndWait` in one centralized call.
As a result, each individual controllers should simply wait until informers have synced, then run the queue to start processing things.

A queue is used to give a few properties:

	if err != nil {
		return probeSet, err
	}
	probeSet.Flush()
	return probeSet, nil
}

func (s *Server) Start() {
	log.Info("CNI ambient server starting")
	s.kubeClient.RunAndWait(s.ctx.Done())
	log.Info("CNI ambient server kubeclient started")
	pods := s.handlers.GetActiveAmbientPodSnapshot()
	err := s.dataplane.ConstructInitialSnapshot(pods)
	if err != nil {