endpoints configTypeURL = "type.googleapis.com/envoy.admin.v3.EndpointsConfigDump"
	clusters  configTypeURL = "type.googleapis.com/envoy.admin.v3.ClustersConfigDump"
	routes    configTypeURL = "type.googleapis.com/envoy.admin.v3.RoutesConfigDump"
	secrets   configTypeURL = "type.googleapis.com/envoy.admin.v3.SecretsConfigDump"
	ecds      configTypeURL = "type.googleapis.com/envoy.admin.v3.EcdsConfigDump"

[entryPoints]
  [entryPoints.http]
    address = ":9999"

[providers]
  [providers.file]
    filename = "routes.toml"
```

这个文件把 Traefik 监听端口设置为 `9999`，并设置要使用另一个文件 `routes.toml`。

!!! tip "提示"

    使用端口 9999 代替标准的 HTTP 端口 80，这样就不必使用管理员权限运行（`sudo`）。

接下来，创建 `routes.toml`：

```TOML hl_lines="5  12  20"
[http]
  [http.middlewares]

    [http.middlewares.api-stripprefix.stripPrefix]

이것은 `0.45.0` 버전과 같거나 높으면서 `0.46.0` 버전 보다는 낮은 버전을 사용했다는 것을 의미합니다. 예를 들어, `0.45.2` 버전과 같은 경우는 해당 조건을 만족합니다.

만약에 Poetry, Pipenv, 또는 그밖의 다양한 설치 도구를 사용한다면, 패키지에 구체적인 버전을 정의할 수 있는 방법을 가지고 있을 것입니다.

## 이용가능한 버전들

[Release Notes](../release-notes.md){.internal-link target=_blank}를 통해 사용할 수 있는 버전들을 확인할 수 있습니다.(예를 들어, 가장 최신의 버전을 확인할 수 있습니다.)


## 버전들에 대해

유의적 버전 관습을 따라서, `1.0.0` 이하의 모든 버전들은 잠재적으로 급변할 수 있습니다.

Für `Item-Input` ist `description` **nicht erforderlich**, es hat kein rotes Sternchen.

Aber für `Item-Output` ist `description` **erforderlich**, es hat ein rotes Sternchen.

<div class="screenshot">
<img src="/img/tutorial/separate-openapi-schemas/image04.png">
</div>

and other connections will be blocked while the large payload is being sent.

## Handlers & Routes

Handlers have a predefined Handler ID. 
In addition, there can be several *static* subroutes used to differentiate between different handlers of the same ID.
A subroute on a client must match a subroute on the server. So routes cannot be used for dynamic routing, unlike HTTP.

- [ ] Developer Infrastructure

**Please check any characteristics that apply to this pull request.**

Pour parler de tâches "synchrones" (en opposition à "asynchrones"), on utilise souvent le terme "séquentiel", car l'ordinateur / le programme va effectuer toutes les étapes d'une tâche séquentiellement avant de passer à une autre tâche, même si ces étapes impliquent de l'attente.

### Concurrence et Burgers

    "license": "Apache-2.0",
    "title": "TensorFlow",
    "upload_type": "software",
    "creators": [
        {
            "name": "TensorFlow Developers"
        }
    ],
    "access_right": "open",
    "notes": "Specific TensorFlow versions can be found in the \"Versions\" list on the right side of this page.<br>See the full list of authors <a href=\"https://github.com/tensorflow/tensorflow/graphs/contributors\">on GitHub</a>."

    response = client.get("/users/")
    assert response.status_code == 200, response.text
    assert response.json() == ["Rick", "Morty"]


def test_dummy_webhook():
    # Just for coverage
    app.webhooks.routes[0].endpoint({})


def test_openapi_schema():
    response = client.get("/openapi.json")
    assert response.status_code == 200, response.text
    assert response.json() == {
        "openapi": "3.1.0",

	// returned early by any other middleware (but after the middleware that
	// sets the amz request id).
	httpTracerMiddleware,
	// Auth middleware verifies incoming authorization headers and routes them
	// accordingly. Client receives a HTTP error for invalid/unsupported
	// signatures.
	//
	// Validates all incoming requests to have a valid date header.
	setAuthMiddleware,