Skip to content

n8n

Open-source workflow automation platform with a visual editor. Build complex AI-powered workflows by connecting LLMs, APIs, databases, and services with a no-code/low-code interface.

Category workflow-automation
Official Docs n8n Documentation
CLI Install ./cli workflow-automation n8n install
CLI Uninstall ./cli workflow-automation n8n uninstall
Namespace n8n

Overview

n8n enables visual workflow automation: - Visual Editor: Drag-and-drop workflow builder - 400+ Integrations: Connect to databases, APIs, SaaS tools - AI Nodes: Built-in LLM, chat, and AI agent nodes - Code Nodes: Write custom JavaScript/Python when needed - Webhooks: Trigger workflows from HTTP requests - Scheduling: Cron-based workflow execution - Self-Hosted: Full control over data and execution

Installation

./cli workflow-automation n8n install

The installer: 1. Renders Helm values with domain configuration 2. Deploys n8n via OCI Helm chart 3. Creates the n8n namespace

No additional environment variables are required (beyond optional DOMAIN).

Verification

# Check pods
kubectl get pods -n n8n

# Check service
kubectl get svc -n n8n

# Port-forward for local access
kubectl port-forward svc/n8n 5678:5678 -n n8n --address 0.0.0.0 &

# Access n8n UI
open http://localhost:5678

Configuration

Environment Variables

Variable Description
DOMAIN Domain for ingress (optional)

Usage

AI Workflows with LiteLLM

Create AI-powered workflows by connecting n8n to LiteLLM:

  1. Add an HTTP Request node or OpenAI node
  2. Set the base URL to http://litellm.litellm:4000/v1
  3. Configure the API key (LiteLLM master key)
  4. Select the model and configure parameters

Example: Chat Workflow

  1. Webhook Trigger: Receive incoming messages
  2. OpenAI Chat: Send to LLM via LiteLLM
  3. HTTP Response: Return the generated response

Example: RAG Workflow

  1. Webhook Trigger: Receive a question
  2. Qdrant Search: Find relevant documents
  3. OpenAI Chat: Generate answer with context
  4. HTTP Response: Return the answer

Troubleshooting

UI not loading

# Check pod status
kubectl get pods -n n8n

# Check logs
kubectl logs -n n8n -l app=n8n

Cannot connect to cluster services

# n8n runs inside the cluster, use internal DNS
# LiteLLM: http://litellm.litellm:4000
# Qdrant: http://qdrant.qdrant:6333
# TEI: http://tei.tei:8080

# Test from n8n pod
kubectl exec -it -n n8n <n8n-pod> -- curl http://litellm.litellm:4000/health

Learn More