Overview

If your application’s runtime isn’t natively supported by a Datadog SDK, you can use the OpenTelemetry SDK to send telemetry to Datadog. This approach gives you traces and metrics without waiting for native runtime support.

This guide walks through instrumenting a Bun application as an example. Bun is compatible with most Node.js APIs, so you can use the OpenTelemetry Node.js SDK to instrument Bun applications. You can apply the same pattern to other unsupported runtimes with OpenTelemetry SDK compatibility.

Support level

Prerequisites

• An OpenTelemetry-compatible backend configured to send data to Datadog. See Send OpenTelemetry Data to Datadog for setup options including the DDOT Collector, OSS Collector with Datadog Exporter, or direct OTLP ingest.
• Bun installed (v1.0 or later).
• A Bun application you want to instrument.

Instrument a Bun application

Install OpenTelemetry packages

From your project’s root directory, install the required OpenTelemetry packages:

bun add @opentelemetry/api \
  @opentelemetry/sdk-node \
  @opentelemetry/sdk-metrics \
  @opentelemetry/auto-instrumentations-node \
  @opentelemetry/exporter-trace-otlp-http \
  @opentelemetry/exporter-metrics-otlp-http \
  @opentelemetry/resources \
  @opentelemetry/semantic-conventions

Create a telemetry initialization module

In Node.js, OpenTelemetry typically loads through the --require flag, which preloads instrumentation before your application code runs. Bun’s module resolution system works differently, so you need to initialize OpenTelemetry programmatically instead.

Create a tracing.ts file that configures the OpenTelemetry Node.js SDK:

import { NodeSDK } from '@opentelemetry/sdk-node';
import { getNodeAutoInstrumentations } from '@opentelemetry/auto-instrumentations-node';
import { OTLPTraceExporter } from '@opentelemetry/exporter-trace-otlp-http';
import { OTLPMetricExporter } from '@opentelemetry/exporter-metrics-otlp-http';
import { PeriodicExportingMetricReader } from '@opentelemetry/sdk-metrics';
import { resourceFromAttributes } from '@opentelemetry/resources';
import { ATTR_SERVICE_NAME, ATTR_SERVICE_VERSION } from '@opentelemetry/semantic-conventions';

const resource = resourceFromAttributes({
  [ATTR_SERVICE_NAME]: process.env.OTEL_SERVICE_NAME || '<YOUR_SERVICE_NAME>',
  [ATTR_SERVICE_VERSION]: process.env.OTEL_SERVICE_VERSION || '1.0.0',
  'telemetry.sdk.runtime': 'bun',
});

const sdk = new NodeSDK({
  resource,
  traceExporter: new OTLPTraceExporter(),
  metricReaders: [new PeriodicExportingMetricReader({
    exporter: new OTLPMetricExporter(),
  })],
  instrumentations: [getNodeAutoInstrumentations({
    // Disable fs instrumentation to avoid compatibility issues with Bun
    '@opentelemetry/instrumentation-fs': { enabled: false },
  })],
});

sdk.start();

process.on('SIGTERM', () => {
  sdk.shutdown()
    .then(() => console.log('OpenTelemetry SDK shut down'))
    .catch((err) => console.error('Error shutting down OpenTelemetry SDK', err))
    .finally(() => process.exit(0));
});

Replace <YOUR_SERVICE_NAME> with your application’s service name.

Initialize telemetry at your application entry point

Import the tracing.ts module before any other application imports. The OpenTelemetry SDK must initialize first to patch libraries for auto-instrumentation.

import './tracing';

// Import your application code after tracing is initialized
import { startApp } from './app';

startApp();

Configure environment variables

Set the following environment variables to configure the OTLP exporter endpoint and your service identity:

export OTEL_EXPORTER_OTLP_ENDPOINT="http://localhost:4318"
export OTEL_SERVICE_NAME="<YOUR_SERVICE_NAME>"

The OTEL_EXPORTER_OTLP_ENDPOINT value depends on your setup:

• Local collector (DDOT or OSS): http://localhost:4318 (default HTTP) or http://localhost:4317 (gRPC)
• Remote collector: Use the collector’s address and port

For additional configuration options, see the OpenTelemetry environment variable specification.

Run your application

Start your Bun application:

bun run index.ts

Add manual instrumentation for Bun-native APIs

Built-in Bun APIs such as Bun.serve(), bun:sqlite, and the native file I/O APIs are not automatically instrumented. To capture telemetry from these APIs, create spans manually with the @opentelemetry/api package.

The following example wraps a Bun.serve() route handler with a custom span:

import { trace, SpanKind } from '@opentelemetry/api';

const tracer = trace.getTracer('bun-app');

export function startServer() {
  Bun.serve({
    port: 3000,
    fetch(req) {
      return tracer.startActiveSpan('handleRequest', { kind: SpanKind.SERVER }, (span) => {
        try {
          span.setAttribute('http.method', req.method);
          span.setAttribute('http.url', req.url);
          return new Response('Hello from Bun!');
        } finally {
          span.end();
        }
      });
    },
  });
}

For more manual instrumentation patterns, see the OpenTelemetry JS instrumentation docs.

Verify traces in Datadog

After your application handles a few requests:

1. Go to APM > Traces in Datadog.
2. Search for your service name.
3. Confirm that traces appear with the expected spans and metadata.

Limitations

• Auto-instrumentation coverage: The OpenTelemetry Node.js auto-instrumentation libraries depend on Bun’s Node.js compatibility layer. Most common libraries (HTTP clients, Express, database drivers) work, but some instrumentation packages may not function as expected.
• Bun-native APIs: Built-in Bun APIs such as Bun.serve(), bun:sqlite, and the native file I/O APIs are not automatically instrumented. See Add manual instrumentation for Bun-native APIs for an example.
• Library compatibility: Not all Node.js instrumentation libraries are guaranteed to work with Bun. Test your specific dependencies and disable any instrumentations that cause errors by passing configuration options to getNodeAutoInstrumentations(). The @opentelemetry/instrumentation-fs package is a common source of issues and is disabled in the example configuration above.

Further reading

추가 유용한 문서, 링크 및 기사:

Datadog and OpenTelemetry CompatibilityDOCUMENTATION Use Custom OpenTelemetry ComponentsDOCUMENTATION Instrument Your Application with OpenTelemetryDOCUMENTATION