Routing OpenTelemetry logs to Sentry using OTLP

If you’ve already instrumented your application with OpenTelemetry, integrating with Sentry doesn’t require replacing your existing instrumentation. By configuring just two environment variables, you can route your logs directly to Sentry’s OTLP endpoint without modifying any SDK code or re-instrumenting your application.

Why You’d Use OTLP Instead of the Native SDK

The primary advantage of OTLP is maintaining vendor-neutral logging code that remains decoupled from any specific observability backend. This approach proves beneficial when you:

• Already have OpenTelemetry logging implemented
• Send logs to multiple observability platforms
• Require vendor-neutral instrumentation
• Work with AI/LLM frameworks using OpenTelemetry natively
• Want to leverage the broader OpenTelemetry ecosystem

However, if starting fresh and using only Sentry, the native Sentry SDK offers superior integration including automatic issue creation, session replay integration, breadcrumb tracking, and error correlation features. OTLP support through Sentry remains in beta and currently lacks these integrated capabilities.

Prerequisites

Required setup includes:

• Active Sentry account (free tier acceptable)
• Node.js 18 or later
• Basic Express.js familiarity

Create a new Sentry project selecting Express as the platform, skipping DSN setup instructions since you’ll use the OTLP endpoint instead.

Getting Your Sentry OTLP Credentials

Navigate to your Sentry project settings:

1. Click Settings in the left sidebar
2. Under Organization, select Projects
3. Choose your project from the list
4. Click Client Keys (DSN) under SDK Setup
5. Select the OpenTelemetry tab
6. Click Expand to reveal OTLP values

Collect these credentials:

• OTLP Logs Endpoint: https://o{ORG_ID}.ingest.us.sentry.io/api/{PROJECT_ID}/integration/otlp/v1/logs
• OTLP Logs Endpoint Headers: x-sentry-auth=sentry sentry_key={YOUR_PUBLIC_KEY}

Most OTLP exporters require headers as key/value pairs rather than full header strings, necessitating header parsing in your application.

Connecting Your OpenTelemetry App to Sentry

Cloning the Starter Application

Execute these commands to clone the payment processing sample app:

git clone https://github.com/getsentry/otlp-logging-sentry.git
cd otlp-logging-sentry
npm install

The starter app includes pre-configured OpenTelemetry SDK, structured logging implementation, multiple log severity levels (INFO, DEBUG, WARN, ERROR), and rich log attributes throughout.

Configuring Sentry as OTLP Destination

Create your environment configuration file:

cp .env.example .env

Edit .env with your Sentry credentials:

OTEL_EXPORTER_OTLP_LOGS_ENDPOINT=https://o{YOUR_ORG_ID}.ingest.us.sentry.io/api/{YOUR_PROJECT_ID}/integration/otlp/v1/logs
OTEL_EXPORTER_OTLP_LOGS_HEADERS=x-sentry-auth=sentry sentry_key={YOUR_PUBLIC_KEY}
OTEL_SERVICE_NAME=payment-processing-service
PORT=3000

Replace placeholders with actual Sentry credentials. The OTEL_SERVICE_NAME enables service-based filtering in Sentry.

Testing the Integration

Start the application:

npm start

Expected console output:

OpenTelemetry logging initialized
Service: payment-processing-service
Payment Processing Service running on http://localhost:3000

Generating Test Logs

In a separate terminal, trigger a payment processing request:

curl -X POST http://localhost:3000/process-payment \
  -H "Content-Type: application/json" \
  -d '{"amount": 100, "currency": "USD"}'

Logs should appear in Sentry within 30-60 seconds.

Troubleshooting

If logs remain absent after verifying credentials, enable OpenTelemetry debug logging:

import { diag, DiagConsoleLogger, DiagLogLevel } from '@opentelemetry/api';
diag.setLogger(new DiagConsoleLogger(), DiagLogLevel.DEBUG);

Consult the OpenTelemetry troubleshooting guide for additional debugging options.

Handling Array Attributes

Array attributes are viewable in log detail but aren’t yet searchable or filterable—a current OTLP beta limitation. For searchable data, use separate attributes or join arrays:

// Can view, but can't search
attributes: {
  'user.preferences': ['fiction', 'mystery', 'sci-fi']
}

// Searchable
attributes: {
  'user.preferences': 'fiction,mystery,sci-fi'
}

Managing Memory Usage

The BatchLogRecordProcessor may be batching excessive logs. Adjust batch sizing:

loggerProvider.addLogRecordProcessor(
  new BatchLogRecordProcessor(logExporter, {
    maxQueueSize: 1000,        // Default: 2048
    maxExportBatchSize: 256,   // Default: 512
    scheduledDelayMillis: 5000 // Default: 1000
  })
);

Consult the BatchLogRecordProcessor documentation for configuration details.