* 还有一个子目录 `app/internal/` 包含另一个 `__init__.py` 文件，因此它是又一个「Python 子包」：`app.internal`。
* `app/internal/admin.py` 是另一个子模块：`app.internal.admin`。

<img src="https://fastapi.tiangolo.com/img/tutorial/bigger-applications/package.drawio.svg">

带有注释的同一文件结构：

```
.
├── app                  # 「app」是一个 Python 包
│   ├── __init__.py      # 这个文件使「app」成为一个 Python 包
│   ├── main.py          # 「main」模块，例如 import app.main

# Aplicaciones más grandes - Múltiples archivos { #bigger-applications-multiple-files }

Si estás construyendo una aplicación o una API web, rara vez podrás poner todo en un solo archivo.

**FastAPI** proporciona una herramienta conveniente para estructurar tu aplicación manteniendo toda la flexibilidad.

/// info | Información

Si vienes de Flask, esto sería el equivalente a los Blueprints de Flask.

///

# Aplicações Maiores - Múltiplos Arquivos { #bigger-applications-multiple-files }

Se você está construindo uma aplicação ou uma API web, é raro que você possa colocar tudo em um único arquivo.

**FastAPI** oferece uma ferramenta conveniente para estruturar sua aplicação, mantendo toda a flexibilidade.

/// info | Informação

Se você vem do Flask, isso seria o equivalente aos Blueprints do Flask.

///

- Ensure Job status updates are batched by 1s. This fixes an unlikely scenario when a sequence of immediately 
  completing pods could trigger a sequence of non-batched Job status updates. ([#118470](https://github.com/kubernetes/kubernetes/pull/118470), [@mimowo](https://github.com/mimowo)) [SIG Apps]

- Added alpha support for asynchronous Pod preemption.
  When the `SchedulerAsyncPreemption` feature gate is enabled, the scheduler now runs API calls to trigger preemptions asynchronously for better performance. ([#128170](https://github.com/kubernetes/kubernetes/pull/128170), [@sanposhiho](https://github.com/sanposhiho)) [SIG Scheduling and Testing]

        .body("abc")
        .onResponseEnd(ShutdownConnection)
        .build(),
    )
    server.enqueue(
      MockResponse(body = "def"),
    )

    // Send a separate request which will trigger a GOAWAY frame on the healthy connection.
    val response = getResponse(newRequest("/"))
    assertContent("abc", response)

    // Ensure the GOAWAY frame has time to be read and processed.
    Thread.sleep(500)

	}
	if !ok { // existing object replication is enabled and object version is unreplicated so far.
		if resetID != "" && oi.ModTime.Before(resetBeforeDate) { // trigger replication if `mc replicate reset` requested
			rd.Replicate = true
			return rd
		}
		// For existing object reset - this condition is needed
		rd.Replicate = tgtStatus == ""
		return rd
	}

- Added the `DetectCacheInconsistency` feature gate, allowing the API server to periodically verify consistency between its `cache` and `etcd`. Detected inconsistencies reported via the `apiserver_storage_consistency_checks_total` metric and trigger purging of affected cache snapshots. ([#132884](https://github.com/kubernetes/kubernetes/pull/132884), [@serathius](https://github.com/serathius)) [SIG API Machinery, Instrumentation and Testing]

    * <a href="https://fastapi.tiangolo.com/tutorial/bigger-applications/" class="external-link" target="_blank">Tutorial - Bigger Applications</a>.

* Fix a bug where soft eviction would not trigger when the threshold was crossed ([#52046](https://github.com/kubernetes/kubernetes/pull/52046), [@dashpole](https://github.com/dashpole))