Datadog APM Trace Injector Plugin

The Datadog APM Trace Injector Plugin automatically injects Datadog APM trace context into your LogLayer logs, enabling correlation between your application logs and distributed traces in Datadog.

Installation

This plugin requires the dd-trace library to be installed in your project.

npm

npm install @loglayer/plugin-datadog-apm-trace-injector dd-trace

yarn

yarn add @loglayer/plugin-datadog-apm-trace-injector dd-trace

pnpm

pnpm add @loglayer/plugin-datadog-apm-trace-injector dd-trace

Configuration

Option	Type	Description
tracerInstance	Tracer	Required. The dd-trace tracer instance
id	string	Optional. Unique identifier for the plugin
disabled	boolean	Optional. Disable the plugin
onError	(error: Error, data?: Record<string, any>) => void	Optional. Error handler for tracer operation failures

Usage

// dd-trace generally needs to be the first import of any project
// as it needs to patch node_module packages before they are imported
import tracer from 'dd-trace';
import { LogLayer } from 'loglayer';
import { datadogTraceInjectorPlugin } from '@loglayer/plugin-datadog-apm-trace-injector';

tracer.init();

// Create the plugin
const traceInjector = datadogTraceInjectorPlugin({
  tracerInstance: tracer,
  // Enable the plugin only if the Datadog API key is set
  enabled: !!process.env.DD_API_KEY
});

// Add to your LogLayer instance
const log = new LogLayer({
  plugins: [traceInjector],
});

// Your logs will now automatically include trace context
log.info('User action completed');

With Error Handling

const traceInjectorWithErrorHandling = datadogTraceInjectorPlugin({
  tracerInstance: tracer,
  onError: (error, data) => {
    console.error('Datadog trace injection failed:', error.message, data);
  },
});

const log = new LogLayer({
  plugins: [traceInjectorWithErrorHandling],
});

Express example

import tracer from 'dd-trace';
import express from 'express';
import { LogLayer, ConsoleTransport } from 'loglayer';
import { datadogTraceInjectorPlugin } from '@loglayer/plugin-datadog-apm-trace-injector';

// Initialize dd-trace
tracer.init();

const app = express();

const log = new LogLayer({
  transport: new ConsoleTransport({
    messageField: 'msg',
    logger: console,
  }),
  plugins: [
    datadogTraceInjectorPlugin({
      tracerInstance: tracer,
    }),
  ],
});

app.get('/', (req, res) => {
  // This log will automatically include trace context
  log.info('Fetching users from database');

  // Your API logic here
  res.json({ users: [] });
});

app.listen(3004, function(err){
    if (err) console.log("Error in server setup")
    console.log("Server listening on Port 3004");
});

Visiting / outputs the following:

{
  dd: {
    trace_id: '689cd152000000002bf3186dadf7c91a',
    span_id: '6991062753777198294',
    service: 'test-service',
    version: '0.0.1'
  },
  msg: 'Fetching users from database'
}

How It Works

The plugin hooks into LogLayer's onBeforeDataOut lifecycle and:

1. Retrieves Active Span: Gets the currently active span from the dd-trace tracer
2. Injects Trace Context: Uses tracer.inject() to add trace and span IDs to the log data
3. Preserves Existing Data: Maintains all existing log data while adding trace context

The injected trace context follows Datadog's log correlation format, allowing you to:

• Correlate logs with traces in the Datadog UI
• Filter logs by trace ID or span ID
• View logs alongside trace spans in distributed tracing views

Trace Context Fields

When a trace is active, the following fields are automatically added to your logs:

• dd.trace_id: The current trace ID
• dd.span_id: The current span ID
• dd.service: The service name (if configured in dd-trace)
• dd.version: The service version (if configured in dd-trace)

Changelog

View the changelog here.