AWS Lambda 高级模式

Lambda 的无服务器模型消除了服务器管理，但需要理解冷启动、内存/CPU 权衡以及架构模式。

冷启动优化

# 错误：在 handler 内部导入（每次冷启动和热调用都执行）
def handler(event, context):
    import boto3  # 慢！
    import json
    client = boto3.client('s3')
    ...

# 正确：顶层导入（仅在冷启动时执行）
import boto3
import json

# 在 handler 外部初始化客户端
s3_client = boto3.client('s3')
dynamodb = boto3.resource('dynamodb')
table = dynamodb.Table('my-table')

def handler(event, context):
    # s3_client 已初始化
    response = s3_client.get_object(Bucket='my-bucket', Key='data.json')
    return json.loads(response['Body'].read())

预置并发

# serverless.yml / CloudFormation
Resources:
  MyFunction:
    Type: AWS::Lambda::Function
    Properties:
      FunctionName: my-api
      Runtime: python3.12
      Handler: index.handler
      Code:
        ZipFile: |
          def handler(event, context):
            return {"statusCode": 200, "body": "Hello"}

  FunctionVersion:
    Type: AWS::Lambda::Version
    Properties:
      FunctionName: !Ref MyFunction

  ProvisionedConcurrency:
    Type: AWS::Lambda::ProvisionedConcurrencyConfig
    Properties:
      FunctionName: !Ref MyFunction
      Qualifier: !GetAtt FunctionVersion.Version
      ProvisionedConcurrentExecutions: 10  # 预预热 10 个实例

Lambda SnapStart（Java）

// 通过初始化后拍摄快照，加快 Java 冷启动
// 在 AWS SAM 模板中：
// SnapStart:
//   ApplyOn: PublishedVersions

@RestController
public class ApiHandler implements RequestHandler<APIGatewayProxyRequestEvent, APIGatewayProxyResponseEvent> {
    
    // 这些在快照期间初始化（而非冷启动）
    private static final ObjectMapper mapper = new ObjectMapper();
    private static final DynamoDbClient dynamoDB = DynamoDbClient.create();
    
    @Override
    public APIGatewayProxyResponseEvent handleRequest(
        APIGatewayProxyRequestEvent input, 
        Context context
    ) {
        // 热路径 - 无需初始化
        String body = processRequest(input.getBody());
        return new APIGatewayProxyResponseEvent()
            .withStatusCode(200)
            .withBody(body);
    }
}

Lambda 层

# 创建包含依赖的层
mkdir -p python/lib/python3.12/site-packages
pip install requests boto3 -t python/lib/python3.12/site-packages/
zip -r my-layer.zip python/

# 发布层
aws lambda publish-layer-version \
  --layer-name my-dependencies \
  --zip-file fileb://my-layer.zip \
  --compatible-runtimes python3.12

# 附加到函数
aws lambda update-function-configuration \
  --function-name my-function \
  --layers arn:aws:lambda:us-east-1:123456789:layer:my-dependencies:1

Lambda 性能调优

# 使用 AWS Lambda Power Tuning 工具（Step Functions）
# https://github.com/alexcasalboni/aws-lambda-power-tuning

# 输入配置
{
  "lambdaARN": "arn:aws:lambda:us-east-1:123:function:my-function",
  "powerValues": [128, 256, 512, 1024, 2048, 3008],
  "num": 50,
  "payload": {"test": "data"},
  "parallelInvocation": true,
  "strategy": "cost"
}

# 结果显示最佳内存/成本/速度权衡
# 通常：512MB 是成本和延迟之间的最佳平衡点

Lambda URL（直接 HTTPS 端点）

// CDK：创建带有 Function URL 的 Lambda
import * as lambda from 'aws-cdk-lib/aws-lambda';
import * as lambdaNodejs from 'aws-cdk-lib/aws-lambda-nodejs';

const fn = new lambdaNodejs.NodejsFunction(this, 'ApiFunction', {
  entry: 'src/handler.ts',
  handler: 'handler',
  runtime: lambda.Runtime.NODEJS_20_X,
  environment: {
    TABLE_NAME: table.tableName,
  },
});

// 添加 Function URL
const fnUrl = fn.addFunctionUrl({
  authType: lambda.FunctionUrlAuthType.NONE,
  cors: {
    allowedOrigins: ['https://myapp.com'],
    allowedMethods: [lambda.HttpMethod.ALL],
    allowedHeaders: ['*'],
  },
});

// Handler
export async function handler(event: any) {
  const method = event.requestContext.http.method;
  const path = event.rawPath;
  
  return {
    statusCode: 200,
    headers: { 'Content-Type': 'application/json' },
    body: JSON.stringify({ path, method }),
  };
}

事件源映射：SQS

// 批量处理 SQS 消息
export async function sqsHandler(event: SQSEvent): Promise<SQSBatchResponse> {
  const failures: SQSBatchItemFailure[] = [];

  await Promise.all(
    event.Records.map(async (record) => {
      try {
        const body = JSON.parse(record.body);
        await processMessage(body);
      } catch (error) {
        console.error(`Failed: ${record.messageId}`, error);
        failures.push({ itemIdentifier: record.messageId });
      }
    })
  );

  // 返回部分失败 - 仅失败的消息重新排队
  return { batchItemFailures: failures };
}

// SQS 触发器的 Lambda 配置
// BatchSize: 10
// MaximumBatchingWindowInSeconds: 30
// FunctionResponseTypes: ReportBatchItemFailures

死信队列模式

Resources:
  ProcessFunction:
    Type: AWS::Lambda::Function
    Properties:
      DeadLetterConfig:
        TargetArn: !GetAtt DLQ.Arn
  
  DLQ:
    Type: AWS::SQS::Queue
    Properties:
      QueueName: process-dlq
      MessageRetentionPeriod: 1209600  # 14 天

  DLQAlarm:
    Type: AWS::CloudWatch::Alarm
    Properties:
      MetricName: ApproximateNumberOfMessagesVisible
      Namespace: AWS/SQS
      Dimensions:
        - Name: QueueName
          Value: process-dlq
      Threshold: 1
      ComparisonOperator: GreaterThanOrEqualToThreshold
      AlarmActions:
        - !Ref AlertTopic

总结

Lambda 优化策略：

• 冷启动：在 handler 外部初始化，Java 使用 SnapStart
• 内存：使用性能调优找到最佳内存设置
• 层：跨函数共享代码/依赖
• 预置并发：为延迟敏感的 API 预预热
• SQS：使用批量项目失败报告实现可靠处理
• DLQ：始终配置死信队列以处理错误