{* ../../docs_src/background_tasks/tutorial002_an_py310.py hl[13,15,22,25] *}

En este ejemplo, los mensajes se escribirán en el archivo `log.txt` *después* de que se envíe el response.

Si hay un query en el request, se escribirá en el log en una tarea en segundo plano.

* Luego devuelve la **response**.

/// note | Detalles Técnicos

Si tienes dependencias con `yield`, el código de salida se ejecutará *después* del middleware.

Si hubiera tareas en segundo plano (cubiertas en la sección [Tareas en segundo plano](background-tasks.md){.internal-link target=_blank}, lo verás más adelante), se ejecutarán *después* de todo el middleware.

///

## Crear un middleware { #create-a-middleware }

* [LastaFlute](https://github.com/lastaflute/lastaflute "LastaFlute"): Framework Web
* [Lasta Job](https://github.com/lastaflute/lasta-job "Lasta Job"): Programador de Tareas
* [Fess Crawler](https://github.com/codelibs/fess-crawler "Fess Crawler"): Rastreador Web

				errs = append(errs, e)
			}
		}
	}
	return tgts, errs
}

// Split targets into two groups:
//
//	group1 contains all targets of type t
//	group2 contains the remaining targets
func splitTargets(targets []Target, t types.TargetType) (group1 []Target, group2 []Target) {
	for _, target := range targets {
		if target.Type() == t {
			group1 = append(group1, target)
		} else {

* Então ele retorna a **resposta**.

/// note | Detalhes Técnicos

Se você tiver dependências com `yield`, o código de saída será executado *depois* do middleware.

Se houver alguma tarefa em segundo plano (abordada na seção [Tarefas em segundo plano](background-tasks.md){.internal-link target=_blank}, que você verá mais adiante), ela será executada *depois* de todo o middleware.

///

## Criar um middleware { #create-a-middleware }

## Adicione a tarefa em segundo plano { #add-the-background-task }

Dentro da sua *função de operação de rota*, passe sua função de tarefa para o objeto de *tarefas em segundo plano* com o método `.add_task()`:

{* ../../docs_src/background_tasks/tutorial001_py39.py hl[14] *}

O `.add_task()` recebe como argumentos:

* Uma função de tarefa a ser executada em segundo plano (`write_notification`).

### Migration and Enforcement

* **New code**: All new contributions must use American English
* **Existing code**: British English spellings in existing code should be updated opportunistically when those areas are modified, unless constrained by backward compatibility
* **Documentation**: Documentation should be updated to use American English, prioritizing user-facing documentation

`SavedModel`) and executes them in parallel on available executors. Running
TensorFlow in a multitenant design mixes the risks described above with the
inherent ones from multitenant configurations. The primary areas of concern are
tenant isolation, resource allocation, model sharing and hardware attacks.

### Tenant isolation

Since any tenants or users providing models, graphs or checkpoints can execute

# Start writeback at 3% memory
vm.dirty_background_ratio=3
# Force writeback at 10% memory
vm.dirty_ratio=10

# Quite a few memory map
# areas may be consumed
vm.max_map_count=524288

# Default is 500000 = 0.5ms
kernel.sched_migration_cost_ns=5000000

# stalled hdd io threads
kernel.hung_task_timeout_secs=85

# network tuning for bigger throughput
net.core.netdev_max_backlog=250000
net.core.somaxconn=16384

	}

	list.RLock()
	defer list.RUnlock()

	targets := make([]Target, 0, len(list.targets))
	for _, tgt := range list.targets {
		targets = append(targets, tgt)
	}

	return targets
}

// Empty returns true if targetList is empty.
func (list *TargetList) Empty() bool {
	list.RLock()
	defer list.RUnlock()

	return len(list.targets) == 0
}

// List - returns available target IDs.