DatadogExporter

将追踪数据发送到 Datadog 的 LLM 可观察性产品，用于监控和分析。

🌐 Sends Tracing data to Datadog's LLM Observability product for monitoring and analytics.

构造函数

🌐 Constructor

new DatadogExporter(config: DatadogExporterConfig)

DatadogExporterConfig

interface DatadogExporterConfig extends BaseExporterConfig {
  apiKey?: string;
  mlApp?: string;
  site?: string;
  service?: string;
  env?: string;
  agentless?: boolean;
  integrationsEnabled?: boolean;
}

扩展自 BaseExporterConfig，包括：

🌐 Extends BaseExporterConfig, which includes:

• logger?: IMastraLogger - 记录器实例
• logLevel?: LogLevel | 'debug' | 'info' | 'warn' | 'error' - 日志级别（默认：INFO）

方法

🌐 Methods

exportTracingEvent

async exportTracingEvent(event: TracingEvent): Promise<void>

将跟踪事件导出到 Datadog LLM 可观测性。

🌐 Exports a tracing event to Datadog LLM Observability.

flush

async flush(): Promise<void>

强制将任何缓冲的跨度刷新到 Datadog，而不关闭导出器。在无服务器环境中非常有用，可以确保在运行时终止之前导出跨度。

🌐 Force flushes any buffered spans to Datadog without shutting down the exporter. Useful in serverless environments where you need to ensure spans are exported before the runtime terminates.

shutdown

async shutdown(): Promise<void>

刷新挂起的数据并关闭导出器。取消任何挂起的清理计时器并禁用 LLM 可观测性。

🌐 Flushes pending data and shuts down the exporter. Cancels any pending cleanup timers and disables LLM Observability.

用法

🌐 Usage

零配置（使用环境变量）

🌐 Zero-Config (using environment variables)

import { DatadogExporter } from "@mastra/datadog";

// Reads from DD_LLMOBS_ML_APP, DD_API_KEY, DD_SITE, DD_ENV
const exporter = new DatadogExporter();

显式配置

🌐 Explicit Configuration

import { DatadogExporter } from "@mastra/datadog";

const exporter = new DatadogExporter({
  mlApp: "my-llm-app",
  apiKey: process.env.DD_API_KEY,
  site: "datadoghq.com",
  env: "production",
});

使用本地 Datadog 代理

🌐 With Local Datadog Agent

const exporter = new DatadogExporter({
  mlApp: "my-llm-app",
  agentless: false, // Use local Datadog Agent
  env: "production",
});

跨度映射

🌐 Span Mapping

Mastra跨度类型映射到Datadog LLMObs跨度种类：

🌐 Mastra span types are mapped to Datadog LLMObs span kinds:

Mastra Span 类型	Datadog 类型
AGENT_RUN	agent
MODEL_GENERATION	workflow
MODEL_STEP	llm
TOOL_CALL	tool
MCP_TOOL_CALL	tool
WORKFLOW_RUN	workflow
所有其他类型	task

所有未映射的跨度类型（包括 MODEL_CHUNK、WORKFLOW_STEP、GENERIC 以及未来的跨度类型）都会自动默认为 task。

🌐 All unmapped span types (including MODEL_CHUNK, WORKFLOW_STEP, GENERIC, and future span types) automatically default to task.

环境变量

🌐 Environment Variables

导出器从这些环境变量中读取配置：

🌐 The exporter reads configuration from these environment variables:

变量	描述
DD_API_KEY	Datadog API 密钥
DD_LLMOBS_ML_APP	ML 应用名称
DD_SITE	Datadog 站点
DD_ENV	环境名称
DD_LLMOBS_AGENTLESS_ENABLED	设置为 'false' 或 '0' 以使用本地代理