## Writing Controllers

The `controllers` package offers a variety of helpers for writing controllers.
These operate in a similar manner as [`controller-runtime`](https://github.com/kubernetes-sigs/controller-runtime) but are *far* smaller and less abstract.

With a few exceptions, Istio controllers typically are split in two phases: construction and running.

### Controllers overview

Below provides a high level overview of controllers in Istiod. For more information about each controller, consulting the controllers Go docs is recommended.

```mermaid
graph BT
    crd("CRD Watcher")
    subgraph Service Discovery
        ksd("Kubernetes Controller")
        sesd("Service Entry Controller")

// IngressClassSpec provides information about the class of an Ingress.
message IngressClassSpec {
  // controller refers to the name of the controller that should handle this
  // class. This allows for different "flavors" that are controlled by the
  // same controller. For example, you may have different parameters for the
  // same implementing controller. This should be specified as a
  // domain-prefixed path no more than 250 characters in length, e.g.

		FieldSelector: strings.Join(fieldSelectors, ","),
	})
	c.queue = controllers.NewQueue("repair pods",
		controllers.WithReconciler(c.Reconcile),
		controllers.WithMaxAttempts(5))
	c.pods.AddEventHandler(controllers.ObjectHandler(c.queue.AddObject))

	return c, nil
}

func (c *Controller) Run(stop <-chan struct{}) {
	kube.WaitForCacheSync("repair controller", stop, c.pods.HasSynced)
	c.queue.Run(stop)
	c.pods.ShutdownHandlers()

[registry.k8s.io/kube-controller-manager:v1.27.11](https://console.cloud.google.com/gcr/images/k8s-artifacts-prod/us/kube-controller-manager) | [amd64](https://console.cloud.google.com/gcr/images/k8s-artifacts-prod/us/kube-controller-manager-amd64), [arm64](https://console.cloud.google.com/gcr/images/k8s-artifacts-prod/us/kube-controller-manager-arm64), [ppc64le](https://console.cloud.google.com/gcr/images/k8s-artifacts-prod/us...

  // where "pod-specific-string" is managed by the StatefulSet controller.
  optional string serviceName = 5;

  // podManagementPolicy controls how pods are created during initial scale up,
  // when replacing pods on nodes, or when scaling down. The default policy is
  // `OrderedReady`, where pods are created in increasing order (pod-0, then
  // pod-1, etc) and the controller will wait until each pod is ready before

  //
  // More completion modes can be added in the future.
  // If the Job controller observes a mode that it doesn't recognize, which
  // is possible during upgrades due to version skew, the controller
  // skips updates for the Job.
  // +optional
  optional string completionMode = 9;

  // suspend specifies whether the Job controller should create Pods or not. If

- The `horizontalpodautoscaling` and `clusterrole-aggregation` controllers now assume the `autoscaling/v1` and `rbac.authorization.k8s.io/v1` APIs are available. If you disable those APIs and do not want to run those controllers, exclude them by passing `--controllers=-horizontalpodautoscaling` or `--controllers=-clusterrole-aggregation` to `kube-controller-manager`. ([#117977](https://github.com/kubernetes/kubernetes/pull/117977), [@liggitt](https://github.com/liggitt))...

            q.setUrl("http://example.com");
        }

        final HostIntervalController controller = new HostIntervalController();
        controller.delayMillisBeforeProcessing = waittime;
        controller.delayMillisAfterProcessing = 0L;
        controller.delayMillisForWaitingNewUrl = 0L;
        controller.delayMillisAtNoUrlInQueue = 0L;

        final Callable<Integer> testCallable = new Callable<Integer>() {

	if !cfg.Enabled {
		repairLog.Info("CNI repair controller is disabled")
		return
	}
	repairLog.Info("starting CNI sidecar repair controller")

	client, err := clientSetup()
	if err != nil {
		repairLog.Fatalf("CNI repair could not construct clientSet: %s", err)
	}

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