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.

      )

    client =
      clientTestRule
        .newClientBuilder()
        .eventListenerFactory(clientTestRule.wrap(eventRecorder))
        .connectTimeout(60, TimeUnit.SECONDS) // Deliberately exacerbate slow fallbacks.
        .dns { dnsResults }
        .fastFallback(true)
        .build()
    url =
      serverIpv4
        .url("/")
        .newBuilder()
        .host("localhost")
        .build()

and then dipped suddenly down, so suddenly that Alice had not a
moment to think about stopping herself before she found herself
falling down a very deep well.

  Either the well was very deep, or she fell very slowly, for she
had plenty of time as she went down to look about her and to
wonder what was going to happen next.  First, she tried to look
down and make out what she was coming to, but it was too dark to

   * bash, it caught on in a flash He did the bash, he did the future bash
   */

  public void testFutureBash() {
    if (isWindows()) {
      return; // TODO: b/136041958 - Running very slowly on Windows CI.
    }
    CyclicBarrier barrier =
        new CyclicBarrier(
            6 // for the setter threads
                + 50 // for the listeners
                + 50 // for the blocking get threads,

completion of replication, `x-minio-internal-replication-reset-arn:<arn>` is set in the metadata with the timestamp of replication and ResetID. For saving iops, the objects which are re-replicated are not first set to `PENDING` state.

This is a slower operation that does not use replication queues and is designed to walk the namespace and replicate objects one at a time so as not to impede server load. Ideally, resync should not be initiated for multiple buckets simultaneously - progress of the...

    try {
      // if waiting works, this should get stuck
      Queues.drain(q, new ArrayList<>(), 1, MAX_VALUE, NANOSECONDS);
      fail();
    } catch (InterruptedException expected) {
      // we indeed waited; a slow thread had enough time to interrupt us
    }
  }

  // same as above; uninterruptible version
  @SuppressWarnings("ThreadPriorityCheck") // TODO: b/175898629 - Consider onSpinWait.

   * this address.
   *
   * If [planToReplace] is non-null, this will swap it for a pooled connection if that pooled
   * connection uses HTTP/2. That results in fewer sockets overall and thus fewer TCP slow starts.
   */
  internal fun planReusePooledConnection(
    planToReplace: ConnectPlan? = null,
    routes: List<Route>? = null,
  ): ReusePlan? {
    val result =

		list.eventsSkipped.Add(int64(len(list.queue)))
		return
	default:
		list.eventsSkipped.Add(1)
		err := fmt.Errorf("concurrent target notifications exceeded %d, configured notification target is too slow to accept events for the incoming request rate", maxConcurrentAsyncSend)
		for id := range targetIDset {
			reqInfo := &logger.ReqInfo{}
			reqInfo.AppendTags("targetID", id.String())

// and returns nil.
func osMkdirAll(dirPath string, perm os.FileMode, baseDir string) error {
	if baseDir != "" {
		if strings.HasPrefix(baseDir, dirPath) {
			return nil
		}
	}

	// Slow path: make sure parent exists and then call Mkdir for path.
	i := len(dirPath)
	for i > 0 && os.IsPathSeparator(dirPath[i-1]) { // Skip trailing path separator.
		i--
	}

	j := i

         * problems can exist with methods like FutureTask.done(), not to mention slow calls to
         * Thread.interrupt() (as discussed in InterruptibleTask). At the end of the day, it's
         * unlikely that cancel() will be slow, so we can probably get away with calling it while
         * holding a lock. Still, it would be nice to avoid somehow.
         */