- Adds a /flagz endpoint for kube-scheduler endpoint ([#128818](https://github.com/kubernetes/kubernetes/pull/128818), [@yongruilin](https://github.com/yongruilin)) [SIG Architecture, Instrumentation, Scheduling and Testing]

E todas elas usarão a mesma conexão WebSocket.

## Usando `Depends` e outros { #using-depends-and-others }

Nos endpoints WebSocket você pode importar do `fastapi` e usar:

* `Depends`
* `Security`
* `Cookie`
* `Header`
* `Path`
* `Query`

Eles funcionam da mesma forma que para outros endpoints FastAPI/*operações de rota*:

{* ../../docs_src/websockets/tutorial002_an_py310.py hl[68:69,82] *}

Crea una variable `ALGORITHM` con el algoritmo usado para firmar el token JWT y configúralo a `"HS256"`.

Crea una variable para la expiración del token.

Define un Modelo de Pydantic que se usará en el endpoint de token para el response.

Crea una función de utilidad para generar un nuevo token de acceso.

{* ../../docs_src/security/tutorial004_an_py310.py hl[4,7,13:15,29:31,79:87] *}

	// Holds the host that was passed using --address
	globalMinioHost = ""
	// Holds the host that was passed using --console-address
	globalMinioConsoleHost = ""

	// Holds the possible host endpoint.
	globalMinioEndpoint    = ""
	globalMinioEndpointURL *xnet.URL

	// globalConfigSys server config system.
	globalConfigSys *ConfigSys

	globalNotificationSys *NotificationSys

	}
	wg.Wait()
}

// Initialize storage disks for each endpoint.
// Errors are returned for each endpoint with matching index.
func initStorageDisksWithErrors(endpoints Endpoints, opts storageOpts) ([]StorageAPI, []error) {
	// Bootstrap disks.
	storageDisks := make([]StorageAPI, len(endpoints))
	g := errgroup.WithNErrs(len(endpoints))
	for index := range endpoints {
		g.Go(func() (err error) {

		if bucket == "" {
			h.ServeHTTP(w, r)
			return
		}

		// MakeBucket requests should be handled at current endpoint
		if r.Method == http.MethodPut && bucket != "" && object == "" && r.URL.RawQuery == "" {
			h.ServeHTTP(w, r)
			return
		}

		// CopyObject requests should be handled at current endpoint as path style
		// requests have target bucket and object in URI and source details are in
		// header fields

			defer lambdaServer.Close()

			lambdaARN := "arn:minio:s3-object-lambda::lambda1:webhook"

			cfg := config.New()
			cfg[config.LambdaWebhookSubSys] = map[string]config.KVS{
				functionID: {
					{Key: "endpoint", Value: lambdaServer.URL},
					{Key: "enable", Value: config.EnableOn},
				},
			}
			cfg[config.APISubSys] = map[string]config.KVS{
				"api": {
					{Key: "gzip", Value: config.EnableOff},
				},

* O frontend armazena esse token temporariamente em algum lugar.
* O usuário clica no frontend para ir para outra seção do aplicativo web.
* O frontend precisa buscar mais dados da API.
    * Mas precisa de autenticação para aquele endpoint específico.
    * Então, para autenticar com nossa API, ele envia um header `Authorization` com o valor `Bearer ` mais o token.
    * Se o token contém `foobar`, o conteúdo do header `Authorization` seria: `Bearer foobar`.

* El usuario hace clic en el frontend para ir a otra sección de la aplicación web frontend.
* El frontend necesita obtener más datos de la API.
    * Pero necesita autenticación para ese endpoint específico.
    * Así que, para autenticarse con nuestra API, envía un `header` `Authorization` con un valor de `Bearer ` más el token.

	exit
fi

# Compare the difference of the list of disks and their location, with the below expected output
diff <(./mc admin info minio1 --json | jq -r '.info.servers[].drives[] | "\(.pool_index),\(.set_index),\(.disk_index) \(.endpoint)"' | sort) <(
	cat <<EOF
0,0,0 http://localhost:9001/tmp/xl/node9001/mnt/disk1
0,0,1 http://localhost:9002/tmp/xl/node9002/mnt/disk1
0,0,2 http://localhost:9001/tmp/xl/node9001/mnt/disk2