}

func (k *KMS) updateMetrics(err error, latency time.Duration) {
	// First, update the latency histogram
	// Therefore, find the first bucket that holds the counter for
	// requests with a latency at least as large as the given request
	// latency and update its and all subsequent counters.
	bucket := slices.IndexFunc(k.latencyBuckets, func(b time.Duration) bool { return latency < b })
	if bucket < 0 {

    @Test
    public void test_constructorWithThrowableCause_OutOfMemoryError() {
        // Test that constructor accepts Error as cause (verifies Throwable parameter change)
        String url = "http://example.com/large-dataset";
        String message = "Crawling failed due to memory exhaustion";
        OutOfMemoryError error = new OutOfMemoryError("Heap space exhausted");

  }

  /**
   * Read 100 MiB of 'a' chars. This was really slow due to a performance bug in [MultipartReader],
   * and will be really slow if we regress the fix for that.
   */
  @Test
  fun `reading a large part with small byteCount`() {
    val multipartBody =
      MultipartBody
        .Builder("foo")
        .addPart(
          headersOf("header-name", "header-value"),
          object : RequestBody() {

    InputStream in = newTestStream(0);
    byte[] b = ByteStreams.toByteArray(in);
    assertThat(b).isEmpty();
  }

  public void testToByteArray_largeStream() throws IOException {
    // well, large enough to require multiple buffers
    byte[] expected = newPreFilledByteArray(10000000);
    InputStream in = new ByteArrayInputStream(expected);
    byte[] b = ByteStreams.toByteArray(in);

      class: mermaid
      format: !!python/name:pymdownx.superfences.fence_code_format ''
  pymdownx.tilde: null
  pymdownx.blocks.admonition:
    types:
    - note
    - attention
    - caution
    - danger
    - error
    - tip
    - hint
    - warning
    - info
    - check
  pymdownx.blocks.details: null
  pymdownx.blocks.tab:
    alternate_style: true
  mdx_include: null

# 更多模型 { #extra-models }

书接上文，多个关联模型这种情况很常见。

特别是用户模型，因为：

* **输入模型**应该含密码
* **输出模型**不应含密码
* **数据库模型**可能需要包含哈希后的密码

/// danger | 危险

不要存储用户的明文密码。始终只存储之后可用于校验的“安全哈希”。

如果你还不了解，可以在[安全性章节](security/simple-oauth2.md#password-hashing)中学习什么是“密码哈希”。

///

## 多个模型 { #multiple-models }

下面的代码展示了不同模型处理密码字段的方式，及使用位置的大致思路：

{* ../../docs_src/extra_models/tutorial001_py310.py hl[7,9,14,20,22,27:28,31:33,38:39] *}

Download a file, print its headers, and print its response body as a string.

The `string()` method on response body is convenient and efficient for small documents. But if the response body is large (greater than 1 MiB), avoid `string()` because it will load the entire document into memory. In that case, prefer to process the body as a stream.

=== ":material-language-kotlin: Kotlin"
    ```kotlin

	ANDW	$0x6006000060060, R27, R5                        // ERROR "cannot use REGTMP as source"
	STP	(R3, R4), 0x1234567(R27)                         // ERROR "REGTMP used in large offset store"
	LDP	0x1234567(R27), (R3, R4)                         // ERROR "REGTMP used in large offset load"
	STP	(R26, R27), 700(R2)                              // ERROR "cannot use REGTMP as source"

* Das **Eingabemodell** muss ein Passwort enthalten können.
* Das **Ausgabemodell** sollte kein Passwort haben.
* Das **Datenbankmodell** müsste wahrscheinlich ein gehashtes Passwort haben.

/// danger | Gefahr

Speichern Sie niemals das Klartextpasswort eines Benutzers. Speichern Sie immer einen „sicheren Hash“, den Sie dann verifizieren können.

* [NGINX Unit](https://unit.nginx.org/howto/fastapi/) : NGINX Unit est un environnement d'exécution d'applications web léger et polyvalent.

## Machine serveur et programme serveur { #server-machine-and-server-program }

Il y a un petit détail sur les noms à garder à l'esprit. 💡