* that once the Thread returns to the Executor that this Executor composes, the interruption
     * will still be present. If the composed Executor is an ExecutorService, it can respond to
     * shutdown() by returning tasks queued on that Thread after {@link #worker} drains the queue.
     */
    @SuppressWarnings("CatchingUnchecked") // sneaky checked exception
    private void workOnQueue() {

<img src="/img/tutorial/websockets/image01.png">

You can type messages in the input box, and send them:

<img src="/img/tutorial/websockets/image02.png">

And your **FastAPI** application with WebSockets will respond back:

<img src="/img/tutorial/websockets/image03.png">

You can send (and receive) many messages:

<img src="/img/tutorial/websockets/image04.png">

And all of them will use the same WebSocket connection.

   *
   * To differentiate the two cases we ping the server when a stream times out. If the overall
   * connection is fine the ping will receive a pong; otherwise it won't.
   *
   * The deadline to respond to this ping attempts to limit the cost of being wrong. If it is too
   * long, streams created while we await the pong will reuse broken connections and inevitably

If it doesn't see an `Authorization` header, or the value doesn't have a `Bearer ` token, it will respond with a 401 status code error (`UNAUTHORIZED`) directly.

You don't even have to check if the token exists to return an error. You can be sure that if your function is executed, it will have a `str` in that token.

		// generally all implementations should simply return
		// 403, but in situations where there is a dependency
		// with the caller to take the client offline purpose
		// fully it should make sure to respond with '412'
		// instead, see cmd/storage-rest-server.go for ideas.
		if c.HealthCheckFn != nil && resp.StatusCode == http.StatusPreconditionFailed {

      // 3. Server tries to send response on the canceled stream
      // Otherwise, there is no guarantee for the sequence. For example, the server may use the
      // first mocked response to respond to the second request.
      val response = super.dispatch(request)
      responseDequeuedLatches[responseIndex].countDown()
      requestCanceledLatches[responseIndex].await()
      responseIndex++

		return matches[1]
	}
	return "unknown"
}

func methodNotAllowedHandler(api string) func(w http.ResponseWriter, r *http.Request) {
	return errorResponseHandler
}

// If none of the http routes match respond with appropriate errors
func errorResponseHandler(w http.ResponseWriter, r *http.Request) {
	if r.Method == http.MethodOptions {
		return
	}

  // number of pods.  Ergo, metrics used must decrease as the pod count is
  // increased, and vice-versa.  See the individual metric source types for
  // more information about how each type of metric must respond.
  // If not set, the default metric will be set to 80% average CPU utilization.
  // +listType=atomic
  // +optional
  repeated MetricSpec metrics = 4;

  // behavior configures the scaling behavior of the target

  // number of pods.  Ergo, metrics used must decrease as the pod count is
  // increased, and vice-versa.  See the individual metric source types for
  // more information about how each type of metric must respond.
  // If not set, the default metric will be set to 80% average CPU utilization.
  // +optional
  repeated MetricSpec metrics = 4;

  // behavior configures the scaling behavior of the target

  // number of pods.  Ergo, metrics used must decrease as the pod count is
  // increased, and vice-versa.  See the individual metric source types for
  // more information about how each type of metric must respond.
  // +optional
  repeated MetricSpec metrics = 4;
}

// HorizontalPodAutoscalerStatus describes the current status of a horizontal pod autoscaler.
message HorizontalPodAutoscalerStatus {