MOVWU	(R5)(R4.UXTW<<3), R10                            // ERROR "invalid index shift amount"
	MOVWU	(R5)(R4<<1), R10                                 // ERROR "invalid index shift amount"
	MOVB	(R5)(R4.SXTW<<5), R10                            // ERROR "invalid index shift amount"
	MOVH	R5, (R6)(R2<<3)                                  // ERROR "invalid index shift amount"

  }

  private fun updateMaxConcurrentStreams(
    connection: Http2Connection,
    amount: Int,
  ) {
    val settings = Settings()
    settings[Settings.MAX_CONCURRENT_STREAMS] = amount
    connection.readerRunnable.applyAndAckSettings(true, settings)
    assertThat(connection.peerSettings[Settings.MAX_CONCURRENT_STREAMS]).isEqualTo(amount)
    taskFaker.runTasks()
  }

      return state.policy.backoffDelayMillis.jitterBy(state.policy.backoffJitterMillis) * 1_000_000
    }
  }

  private fun Long.jitterBy(amount: Int): Long {
    return this + ThreadLocalRandom.current().nextInt(amount * -1, amount)
  }

  class AddressState(
    val address: Address,
    val queue: TaskQueue,
    var policy: ConnectionPool.AddressPolicy,
  ) {
    /**

  }

  /**
   * Sends the given method call to this thread.
   *
   * @throws TimeoutException if this thread does not accept the request within a reasonable amount
   *     of time
   */
  private void sendRequest(String methodName, Object... arguments) throws Exception {
    if (!requestQueue.offer(new Request(methodName, arguments), TIMEOUT_MILLIS, MILLISECONDS)) {

If you have an API that does a comparable amount of computations every time and you have a lot of clients, then the **CPU utilization** will probably *also be stable* (instead of constantly going up and down quickly).

### Examples of Replication Tools and Strategies

There can be several approaches to achieve this, and I'll tell you more about specific strategies in the next chapters, for example when talking about Docker and containers.

   * "expected arrival time of the next request" is actually in the past, then the difference (now -
   * past) is the amount of time that the RateLimiter was formally unused, and it is that amount of
   * time which we translate to storedPermits. (We increase storedPermits with the amount of permits
   * that would have been produced in that idle time). So, if rate == 1 permit per second, and

		tok := p.next()
		ext = tok.String()
	}
	if p.peek() == lex.LSH {
		// parses left shift amount applied after extension: <<Amount
		p.get(lex.LSH)
		tok := p.get(scanner.Int)
		amount, err := strconv.ParseInt(tok.String(), 10, 16)
		if err != nil {
			p.errorf("parsing left shift amount: %s", err)
		}
		num = int16(amount)
	} else if p.peek() == '[' {
		// parses an element: [Index]
		p.get('[')

you will have a more or less well-defined, stable, and limited amount of memory consumed by each of those containers (more than one if they are replicated).

And then you can set those same memory limits and requirements in your configurations for your container management system (for example in **Kubernetes**). That way it will be able to **replicate the containers** in the **available machines** taking into account the amount of memory needed by them, and the amount available in the machines...

  }

  /**
   * Sends the given method call to this thread.
   *
   * @throws TimeoutException if this thread does not accept the request within a reasonable amount
   *     of time
   */
  private void sendRequest(String methodName, Object... arguments) throws Exception {
    if (!requestQueue.offer(new Request(methodName, arguments), TIMEOUT_MILLIS, MILLISECONDS)) {

    }
  }

  /**
   * This neat test shows that no matter what weights we use in our requests, if we push X amount of
   * permits in a cool state, where X = rate * timeToCoolDown, and we have specified a
   * timeToWarmUp() period, it will cost as the prescribed amount of time. E.g., calling
   * [acquire(5), acquire(1)] takes exactly the same time as [acquire(2), acquire(3), acquire(1)].
   */