*         <li>and {@link AbstractInvocationHandler#equals} returns true for the {@link
   *             InvocationHandler} of {@code argument}
   *       </ul>
   *   <li>other method calls are dispatched to {@link #handleInvocation}.
   * </ul>
   */
  @Override
  public final @Nullable Object invoke(
      Object proxy, Method method, @Nullable Object @Nullable [] args) throws Throwable {
    if (args == null) {

    this.bus = bus;
    this.target = checkNotNull(target);
    this.method = method;
    method.setAccessible(true);

    this.executor = bus.executor();
  }

  /** Dispatches {@code event} to this subscriber using the proper executor. */
  final void dispatchEvent(Object event) {
    executor.execute(
        () -> {
          try {
            invokeSubscriberMethod(event);

So we have a dedicated thread for every socket that just reads frames and dispatches them.

The reader thread must never run application-layer code. Otherwise one slow stream can hold up the entire connection.

      iptables, saving a bit of memory/time/cpu. (#106373, @aojea) [SIG Network]
- Watch requests that are delegated to aggregated apiservers no longer reserve concurrency units (seats) in the API Priority and Fairness dispatcher for their entire duration. (#105827, @benluddy) [SIG API Machinery]

## Dependencies

### Added
_Nothing has changed._

### Changed
- k8s.io/kube-openapi: 9528897 → 2043435

// ReadWriter stores pointers to a [Reader] and a [Writer].
// It implements [io.ReadWriter].
type ReadWriter struct {
	*Reader
	*Writer
}

// NewReadWriter allocates a new [ReadWriter] that dispatches to r and w.
func NewReadWriter(r *Reader, w *Writer) *ReadWriter {
	return &ReadWriter{r, w}

const maxDurationSecondsForLog = 5

func (sys *IAMSys) periodicRoutines(ctx context.Context, baseInterval time.Duration) {
	// Watch for IAM config changes for iamStorageWatcher.
	watcher, isWatcher := sys.store.IAMStorageAPI.(iamStorageWatcher)
	if isWatcher {
		go func() {
			ch := watcher.watch(ctx, iamConfigPrefix)
			for event := range ch {
				if err := sys.loadWatchedEvent(ctx, event); err != nil {
					// we simply log errors

- Executed API calls dispatched during pod scheduling asynchronously if the `SchedulerAsyncAPICalls` feature gate was enabled.

- Two changes to the `/debug/api_priority_and_fairness/dump_priority_levels` endpoint of API Priority and Fairness: added total number of dispatched, timed-out, rejected and cancelled requests; output now sorted by `PriorityLevelName`. ([#112393](https://github.com/kubernetes/kubernetes/pull/112393), [@borgerli](https://github.com/borgerli))