本文為英文版的機器翻譯版本，如內容有任何歧義或不一致之處，概以英文版為準。

# Node.js 和 DAX
<a name="DAX.client.run-application-nodejs-3"></a>

# Node.js 的預設用戶端組態
<a name="DAX-client-config-JS"></a>

設定 DAX JavaScript SDK 用戶端時，您可以自訂各種參數，以最佳化效能、連線處理和錯誤恢復能力。下表概述控制用戶端與 DAX 叢集互動方式的預設組態設定，包括逾時值、重試機制、憑證管理和運作狀態監控選項。如需詳細資訊，請參閱 [DynamoDBClient 操作](https://docs.aws.amazon.com/AWSJavaScriptSDK/v3/latest/client/dynamodb/)。


**DAX JS SDK 用戶端預設值**  

| 參數 | 類型 | 說明 | 
| --- | --- | --- | 
|  `region` 選擇性  |  `string`  |   AWS 區域 用於 DAX 用戶端的 （範例 - 'us-east-1')。如果未透過環境變數提供，則此為必要參數。  | 
|  `endpoint` 必要  |  `string`  | 軟體開發套件所連線叢集的端點。 範例： 未加密：dax-cluster-name.region.amazonaws.com 加密：daxs://my-cluster.l6fzcv.dax-clusters.us-east-1.amazonaws.com  | 
|  `requestTimeout` 預設 6000 毫秒  |  `number`  | 這用於定義用戶端等待 DAX 回應的時間上限。  | 
|  `writeRetries` 預設 1  |  `number`  | 寫入請求失敗時的重試次數。  | 
|  `readRetries` 預設 1  |  `number`  | 讀取請求失敗時的重試次數。  | 
|  `maxRetries` 預設 1  |  `number`  | 請求失敗時的重試次數上限。 如果已設定 readRetries/writeRetries，則 readRetries 和 writeRetries 中設定的組態會優先於 maxRetries。  | 
|  `connectTimeout` 預設 10000 毫秒  |  `number`  | 建立任何叢集節點連線的逾時值 (以毫秒為單位)。  | 
|  `maxConcurrentConnections` 預設 100  |  `number`  | 限制用戶端執行個體可在 DAX 叢集中為每個節點建立的並行連線總數。  | 
|  `maxRetryDelay` 預設 7000 毫秒  |  `number`  | DAX 伺服器將 `waitForRecoveryBeforeRetrying` 旗標設定為 true 表示需要復原時，用戶端會先暫停然後再重試。在這些復原期間，`maxRetryDelay` 參數會決定重試之間的最長等待時間。此復原專屬組態僅適用於處於復原模式的 DAX 伺服器。對於所有其他情境，重試行為會遵循兩種模式之一：根據重試計數的指數延遲 (由 `writeRetries`、`readRetries`、或 `maxRetries` 參數控管)，或根據例外狀況類型立即重試。  | 
|  `credentials` 選擇性  |  `[AwsCredentialIdentity](https://docs.aws.amazon.com/AWSJavaScriptSDK/v3/latest/Package/-aws-sdk-credential-providers/)` \$1 `[AwsCredentialIdentityProvider](https://docs.aws.amazon.com/AWSJavaScriptSDK/v3/latest/Package/-aws-sdk-credential-providers/)`  |  用於驗證請求的 AWS 登入資料。這能夠以 AwsCredentialIdentity 或 AwsCredentialIdentityProvider 的形式提供。如果未提供， AWS 軟體開發套件會自動使用預設憑證供應者鏈結。範例：`\$1 accessKeyId: 'AKIA...', secretAccessKey: '...', sessionToken: '...' \$1` \$1 @default 使用預設 AWS 憑證供應者鏈結。  | 
|  `healthCheckInterval` 預設 5000 毫秒  |  `number`  | 叢集運作狀態檢查之間的間隔 (以毫秒為單位)。間隔時間越短，檢查頻率就更高。  | 
|  `healthCheckTimeout` 預設 1000 毫秒  |  `number`  | 完成運作狀態檢查的逾時 (以毫秒為單位)。  | 
|  `skipHostnameVerification` 預設 false  |  `boolean`  |  略過 TLS 連線的主機名稱驗證。這不會影響未加密叢集。預設為執行主機名稱驗證，將此設定為 True 會略過驗證。請務必了解將其關閉的意涵，也就是無法驗證您連線的叢集。  | 
|  `unhealthyConsecutiveErrorCount` 預設 5  |  `number`  | 設定在運作狀態檢查間隔內向節點發出運作狀態不佳訊號所需的連續錯誤次數。  | 
|  `clusterUpdateInterval` 預設 4000 毫秒  |  `number`  | 傳回輪詢叢集成員以進行成員資格變更之間的間隔。  | 
|  `clusterUpdateThreshold` 預設 125  |  `number`  | 傳回低於閾值的叢集不會接受輪詢進行成員資格變更。  | 
|  `credentailProvider` 選用 \$1 預設 null  |  `[AwsCredentialIdentityProvider](https://docs.aws.amazon.com/AWSJavaScriptSDK/v3/latest/Package/-aws-sdk-credential-providers/)`  | 用於驗證 DAX 請求之 AWS 登入資料的使用者定義提供者。  | 


**DaxDocument 的分頁組態**  

| 名稱 | 類型 | 詳細資訊 | 
| --- | --- | --- | 
|  `client`  |  DaxDocument  |  DaxDocument 類型的執行個體。  | 
|  `pageSize`  |  number  |  決定每頁的項目數。  | 
|  `startingToken` 選用  |  any  |  先前回應的 LastEvaluatedKey 可用於後續請求。  | 

如需了解分頁的使用方式，請參閱 [TryDax.js](DAX.client.tutorial-TryDax.md)。

# 移轉至 DAX Node.js SDK V3
<a name="DAX.client.run-application-nodejs-3-migrating"></a>

此移轉指南將協助您轉換現有的 DAX Node.js 應用程式。新的 SDK 需要 Node.js 18 或更高版本，並在建構 DynamoDB Accelerator 程式碼方面引入幾項重要變更。本指南將逐步解說其中的主要差異，包括語法變更、新的匯入方法，以及更新的非同步程式設計模式。

## 使用 V2 Node.js DAX
<a name="DAX.client.run-application-nodejs-3-migrating-V2-usage"></a>

```
const AmazonDaxClient = require('amazon-dax-client');
const AWS = require('aws-sdk');

var region = "us-west-2";

AWS.config.update({
  region: region,
});

var client = new AWS.DynamoDB.DocumentClient();

if (process.argv.length > 2) {
  var dax = new AmazonDaxClient({
    endpoints: [process.argv[2]],
    region: region,
  });
  client = new AWS.DynamoDB.DocumentClient({ service: dax });
}

// Make Get Call using Dax
var params = {
    TableName: 'TryDaxTable',
    pk: 1,
    sk: 1
}
client.get(params, function (err, data) {
    if (err) {
        console.error(
            "Unable to read item. Error JSON:",
            JSON.stringify(err, null, 2)
          );
    } else {
        console.log(data);
    }
});
```

## 使用 V3 Node.js DAX
<a name="DAX.client.run-application-nodejs-3-migrating-V3-dax-usage"></a>

使用 DAX 時建議採用 Node.js V3 Node 版本 18 或更高版本。若要移至 Node 18，請使用下列項目：

```
import { DaxDocument } from '@amazon-dax-sdk/lib-dax';
import { DynamoDBDocument } from '@aws-sdk/lib-dynamodb';
import { DynamoDBClient } from '@aws-sdk/client-dynamodb';

let client: DynamoDBDocument | DaxDocument = DynamoDBDocument.from(
  new DynamoDBClient({ region: 'us-west-2' })
);

if (process.argv.length > 2) {
  client = new DaxDocument({
    endpoints: [process.argv[2]],
    region: 'us-west-2',
  });
}

const params = {
  TableName: 'TryDaxTable',
  Key: { pk: 1, sk: 1 },
};

try {
  const results = await client.get(params);
  console.log(results);
} catch (err) {
  console.error(err);
}
```

DAX SDK for Node.js v3.x 相容於 [AWS SDK for Node.js v3.x](https://docs.aws.amazon.com/AWSJavaScriptSDK/v3/latest/introduction/)。DAX SDK for Node.js v3.x 支援使用[彙總](https://docs.aws.amazon.com/AWSJavaScriptSDK/v3/latest/introduction/#high-level-concepts)用戶端。請注意，DAX 不支援建立裸機用戶端。如需不支援功能的詳細資訊，請參閱 [功能與 AWS SDK V3 不相同](#DAX.client.run-application-nodejs-3-not-in-parity)。

請按照此程序操作，在 Amazon EC2 執行個體上執行 Node.js 範例應用程式。

**執行 DAX 的 Node.js 範例**

1. 在您的 Amazon EC2 執行個體上設定 Node.js，如下所示：

   1. 安裝節點版本管理工具 (`nvm`)。

      ```
      curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.40.1/install.sh | bash
      ```

   1. 使用 nvm 安裝 Node.js。

      ```
      nvm install 18
      ```

   1. 使用 nvm 來使用 Node 18

      ```
      nvm use 18
      ```

   1. 測試 Node.js 已安裝且正常運作。

      ```
      node -e "console.log('Running Node.js ' + process.version)"
      ```

      這應該會顯示以下訊息。

      `Running Node.js v18.x.x`

1. 使用節點套件管理工具 (`npm`) 安裝 DaxDocument Node.js 用戶端。

   ```
   npm install @amazon-dax-sdk/lib-dax
   ```

## TryDax 範本程式碼
<a name="DAX.client.run-application-nodejs-3-TryDax-sample-code"></a>

若要評估 DynamoDB Accelerator (DAX) 的效能優勢，請依照下列步驟執行範例測試，以比較標準 DynamoDB 和 DAX 叢集之間的讀取操作時間。

1. 在設定工作空間並將 `lib-dax` 安裝為依存項目之後，請將 [TryDax.js](DAX.client.tutorial-TryDax.md) 複製到您的專案中。

1. 針對 DAX 叢集執行程式。若要判斷您 DAX 叢集的端點，請選擇下列其中一個項目：
   +  **使用 DynamoDB 主控台**：選擇您的 DAX 叢集。叢集端點會在主控台上顯示，如以下範例。

     ```
     dax://my-cluster.l6fzcv.dax-clusters.us-east-1.amazonaws.com
     ```
   + **使用 AWS CLI** ：輸入下列命令。

     ```
     aws dax describe-clusters --query "Clusters[*].ClusterDiscoveryEndpoint"
     ```

     如下列範例所示，叢集端點會在輸出上顯示。

     ```
     {
         "Address": "my-cluster.l6fzcv.dax-clusters.us-east-1.amazonaws.com",
         "Port": 8111,
         "URL": "dax://my-cluster.l6fzcv.dax-clusters.us-east-1.amazonaws.com"
     }
     ```

1. 現在請將叢集端點指定為命令列參數以執行程式。

   ```
   node TryDax.js dax://my-cluster.l6fzcv.dax-clusters.us-east-1.amazonaws.com
   ```

   您應該會看到類似下列的輸出：

   ```
   Attempting to create table; please wait...
   Successfully created table. Table status: ACTIVE
   Writing data to the table...
   Writing 20 items for partition key:  1
   Writing 20 items for partition key:  2
   Writing 20 items for partition key:  3
   ...
   Running GetItem Test
           Total time: 153555.10 µs - Avg time: 383.89 µs
           Total time: 44679.96 µs - Avg time: 111.70 µs
           Total time: 36885.86 µs - Avg time: 92.21 µs
           Total time: 32467.25 µs - Avg time: 81.17 µs
           Total time: 32202.60 µs - Avg time: 80.51 µs
   Running Query Test
           Total time: 14869.25 µs - Avg time: 2973.85 µs
           Total time: 3036.31 µs - Avg time: 607.26 µs
           Total time: 2468.92 µs - Avg time: 493.78 µs
           Total time: 2062.53 µs - Avg time: 412.51 µs
           Total time: 2178.22 µs - Avg time: 435.64 µs
   Running Scan Test
           Total time: 2395.88 µs - Avg time: 479.18 µs
           Total time: 2207.16 µs - Avg time: 441.43 µs
           Total time: 2443.14 µs - Avg time: 488.63 µs
           Total time: 2038.24 µs - Avg time: 407.65 µs
           Total time: 1972.17 µs - Avg time: 394.43 µs
   Running Pagination Test
   Scan Pagination
   [
     { pk: 1, sk: 1, someData: 'XXXXXXXXXX' },
     { pk: 1, sk: 2, someData: 'XXXXXXXXXX' },
     { pk: 1, sk: 3, someData: 'XXXXXXXXXX' }
   ]
   [
     { pk: 1, sk: 4, someData: 'XXXXXXXXXX' },
     { pk: 1, sk: 5, someData: 'XXXXXXXXXX' },
     { pk: 1, sk: 6, someData: 'XXXXXXXXXX' }
   ]
   ...
   Query Pagination
   [
     { pk: 1, sk: 1, someData: 'XXXXXXXXXX' },
     { pk: 1, sk: 2, someData: 'XXXXXXXXXX' },
     { pk: 1, sk: 3, someData: 'XXXXXXXXXX' }
   ]
   [
     { pk: 1, sk: 4, someData: 'XXXXXXXXXX' },
     { pk: 1, sk: 5, someData: 'XXXXXXXXXX' },
     { pk: 1, sk: 6, someData: 'XXXXXXXXXX' }
   ]
   ...
   Attempting to delete table; please wait...
   Successfully deleted table.
   ```

   請記下計時資訊。`GetItem`、`Query`、`Scan` 測試所需的微秒數。

1. 在此情況下，您會針對 DAX 叢集執行程式。現在您將再次執行該程式，但此次針對的是 DynamoDB。

1. 現在重新執行程式，但這一次叢集端點 URL 並沒有作為命令列參數。

   ```
   node TryDax.js
   ```

   查看輸出，並記下計時資訊。使用 DAX 的 `GetItem`、`Query` 和 `Scan` 已耗用時間應遠低於使用 DynamoDB 的已耗用時間。

## 功能與 AWS SDK V3 不相同
<a name="DAX.client.run-application-nodejs-3-not-in-parity"></a>
+ [裸機](https://docs.aws.amazon.com/AWSJavaScriptSDK/v3/latest/introduction/#high-level-concepts)用戶端：Dax Node.js V3 不支援裸機用戶端。

  ```
  const dynamoDBClient = new DynamoDBClient({ region: 'us-west-2' });
  const regularParams = {
      TableName: 'TryDaxTable',
      Key: {
          pk: 1,
          sk: 1
      }
  };
  // The DynamoDB client supports the send operation.
  const dynamoResult = await dynamoDBClient.send(new GetCommand(regularParams));
  
  // However, the DaxDocument client does not support the send operation.
  const daxClient = new DaxDocument({
      endpoints: ['your-dax-endpoint'],
      region: 'us-west-2',
  });
  
  const params = {
      TableName: 'TryDaxTable',
      Key: {
          pk: 1,
          sk: 1
      }
  };
  
  // This will throw an error - send operation is not supported for DAX. Please refer to documentation.
  const result = await daxClient.send(new GetCommand(params));
  console.log(result);
  ```
+ [中介軟體堆疊](https://aws.amazon.com/blogs/developer/middleware-stack-modular-aws-sdk-js/)：Dax Node.js V3 不支援使用中介軟體函式。

  ```
  const dynamoDBClient = new DynamoDBClient({ region: 'us-west-2' });
  // The DynamoDB client supports the middlewareStack.
  dynamoDBClient.middlewareStack.add(
    (next, context) =>> async (args) => {
      console.log("Before operation:", args);
      const result = await next(args);
      console.log("After operation:", result);
      return result;
    },
    {
      step: "initialize", // or "build", "finalizeRequest", "deserialize"
      name: "loggingMiddleware",
    }
  );
  
  // However, the DaxDocument client does not support the middlewareStack.
  const daxClient = new DaxDocument({
      endpoints: ['your-dax-endpoint'],
      region: 'us-west-2',
  });
  
  // This will throw an error - custom middleware and middlewareStacks are not supported for DAX. Please refer to documentation.
  daxClient.middlewareStack.add(
    (next, context) => async (args) => {
      console.log("Before operation:", args);
      const result = await next(args);
      console.log("After operation:", result);
      return result;
    },
    {
      step: "initialize", // or "build", "finalizeRequest", "deserialize"
      name: "loggingMiddleware",
    }
  );
  ```

# TryDax.js
<a name="DAX.client.tutorial-TryDax"></a>

```
import { DynamoDB, waitUntilTableExists, waitUntilTableNotExists } from "@aws-sdk/client-dynamodb";
import { DaxDocument, daxPaginateScan, daxPaginateQuery } from "@amazon-dax-sdk/lib-dax";
import { DynamoDBDocument, paginateQuery, paginateScan } from "@aws-sdk/lib-dynamodb";

const region = "us-east-1";
const tableName = "TryDaxTable";

// Determine the client (DynamoDB or DAX)
let client = DynamoDBDocument.from(new DynamoDB({ region }));
if (process.argv.length > 2) {
  client = new DaxDocument({ region, endpoint: process.argv[2] });
}

// Function to create table
async function createTable() {
  const dynamodb = new DynamoDB({ region });
  const params = {
    TableName: tableName,
    KeySchema: [
      { AttributeName: "pk", KeyType: "HASH" },
      { AttributeName: "sk", KeyType: "RANGE" },
    ],
    AttributeDefinitions: [
      { AttributeName: "pk", AttributeType: "N" },
      { AttributeName: "sk", AttributeType: "N" },
    ],
    ProvisionedThroughput: { ReadCapacityUnits: 25, WriteCapacityUnits: 25 },
  };

  try {
    console.log("Attempting to create table; please wait...");
    await dynamodb.createTable(params);
    await waitUntilTableExists({ client: dynamodb, maxWaitTime: 30 }, { TableName: tableName });
    console.log("Successfully created table. Table status: ACTIVE");
  } catch (err) {
    console.error("Error in table creation:", err);
  }
}

// Function to insert data
async function writeData() {
  console.log("Writing data to the table...");
  const someData = "X".repeat(10);
  for (let ipk = 1; ipk <= 20; ipk++) {
    console.log("Writing 20 items for partition key: ", ipk)
    for (let isk = 1; isk <= 20; isk++) {
      try {
        await client.put({ TableName: tableName, Item: { pk: ipk, sk: isk, someData } });
      } catch (err) {
        console.error("Error inserting data:", err);
      }
    }
  }
}

// Function to test GetItem
async function getItemTest() {
  console.log("Running GetItem Test");
  for (let i = 0; i < 5; i++) {
    const startTime = performance.now();
    const promises = [];
    for (let ipk = 1; ipk <= 20; ipk++) {
      for (let isk = 1; isk <= 20; isk++) {
        promises.push(client.get({ TableName: tableName, Key: { pk: ipk, sk: isk } }));
      }
    }
    await Promise.all(promises);
    const endTime = performance.now();
    const iterTime = (endTime - startTime) * 1000;
    const totalTime = iterTime.toFixed(2).padStart(3, ' ');
    const avgTime = (iterTime / 400).toFixed(2).padStart(3, ' ');
    console.log(`\tTotal time: ${totalTime} \u00B5s - Avg time: ${avgTime} \u00B5s`);
  }
}

// Function to test Query
async function queryTest() {
  console.log("Running Query Test");
  for (let i = 0; i < 5; i++) {
    const startTime = performance.now();
    const promises = [];
    for (let pk = 1; pk <= 5; pk++) {
      const params = {
        TableName: tableName,
        KeyConditionExpression: "pk = :pkval and sk between :skval1 and :skval2",
        ExpressionAttributeValues: { ":pkval": pk, ":skval1": 1, ":skval2": 2 },
      };
      promises.push(client.query(params));
    }
    await Promise.all(promises);
    const endTime = performance.now();
    const iterTime = (endTime - startTime) * 1000;
    const totalTime = iterTime.toFixed(2).padStart(3, ' ');
    const avgTime = (iterTime / 5).toFixed(2).padStart(3, ' ');
    console.log(`\tTotal time: ${totalTime} \u00B5s - Avg time: ${avgTime} \u00B5s`);
  }
}

// Function to test Scan
async function scanTest() {
  console.log("Running Scan Test");
  for (let i = 0; i < 5; i++) {
    const startTime = performance.now();
    const promises = [];
    for (let pk = 1; pk <= 5; pk++) {
      const params = {
        TableName: tableName,
        FilterExpression: "pk = :pkval and sk between :skval1 and :skval2",
        ExpressionAttributeValues: { ":pkval": pk, ":skval1": 1, ":skval2": 2 },
      };
      promises.push(client.scan(params));
    }
    await Promise.all(promises);
    const endTime = performance.now();
    const iterTime = (endTime - startTime) * 1000;
    const totalTime = iterTime.toFixed(2).padStart(3, ' ');
    const avgTime = (iterTime / 5).toFixed(2).padStart(3, ' ');
    console.log(`\tTotal time: ${totalTime} \u00B5s - Avg time: ${avgTime} \u00B5s`);
  }
}

// Function to test Pagination
async function paginationTest() {
  console.log("Running Pagination Test");
  console.log("Scan Pagination");
  const scanParams = { TableName: tableName };
  const paginator = process.argv.length > 2 ? daxPaginateScan : paginateScan;
  for await (const page of paginator({ client, pageSize: 3 }, scanParams)) {
    console.log(page.Items);
  }

  console.log("Query Pagination");
  const queryParams = {
    TableName: tableName,
    KeyConditionExpression: "pk = :pkval and sk between :skval1 and :skval2",
    ExpressionAttributeValues: { ":pkval": 1, ":skval1": 1, ":skval2": 10 },
  };
  const queryPaginator = process.argv.length > 2 ? daxPaginateQuery : paginateQuery;
  for await (const page of queryPaginator({ client, pageSize: 3 }, queryParams)) {
    console.log(page.Items);
  }
}

// Function to delete the table
async function deleteTable() {
  const dynamodb = new DynamoDB({ region });
  console.log("Attempting to delete table; please wait...")
  try {
    await dynamodb.deleteTable({ TableName: tableName });
    await waitUntilTableNotExists({ client: dynamodb, maxWaitTime: 30 }, { TableName: tableName });
    console.log("Successfully deleted table.");
  } catch (err) {
    console.error("Error deleting table:", err);
  }
}

// Execute functions sequentially
(async function () {
  await createTable();
  await writeData();
  await getItemTest();
  await queryTest();
  await scanTest();
  await paginationTest();
  await deleteTable();
})();
```