N8n 2.x.x + OpenTelemetry (Auto-Build Docker Image)

Tips & Tricks
,
1

Hi everyone :waving_hand:

I’ve just published a repository that automatically builds n8n 2.x.x images instrumented with OpenTelemetry, based on the official n8n releases:

:backhand_index_pointing_right: https://github.com/gabrielhmsantos/n8n-tracekit

The project monitors new 2.x.x releases from n8n and automatically builds and publishes a Docker image to Docker Hub, keeping it aligned with the latest stable versions.

This project was started after reading the great community post:

:backhand_index_pointing_right: https://community.n8n.io/t/n8n-successfully-instrumented-with-opentelemetry/78468

Huge thanks to @StuartJohnson and @muxxi, who originally started the OpenTelemetry instrumentation work for n8n 1.x.x.

I’ve adapted, validated and extended the approach to work with n8n 2.x.x, and automated the build pipeline so new versions are compiled automatically as they are released.

What this repository provides

  • :white_check_mark: n8n 2.x.x pre-instrumented with OpenTelemetry

  • :white_check_mark: Automatic build when new n8n releases are published

  • :white_check_mark: Docker images ready to use

  • :white_check_mark: Compatible with OTLP (Tempo, Honeycomb, Datadog, New Relic, etc.)

  • :white_check_mark: Environment-based configuration (OTEL_* variables)

The goal is to make tracing adoption in n8n much easier for production environments.

Open for improvements :rocket:

The repository is fully open for:

  • Improvements in tracing instrumentation

  • Better span naming and attributes

  • Metrics/log correlation

  • Collector integration suggestions

  • Performance optimizations

If you’re running n8n in production with observability, feedback is very welcome.

Let’s make n8n observability first-class :blue_heart:

1 Like
2

Quick update :rocket:

I’ve now added Langfuse integration to the project.

This means you can send traces not only to standard OTLP backends (Tempo, Honeycomb, etc.), but also leverage Langfuse for LLM-specific observability — especially useful if you’re running AI-powered workflows inside n8n.

The setup remains fully environment-based, and Langfuse can run alongside other OTel exporters.

Feedback and improvements are very welcome :raising_hands: