How to Set Up n8n: A Practical Guide

TL;DR

• Self-host n8n in 2 minutes with Docker: docker run -it --rm --name n8n -p 5678:5678 n8nio/n8n
• Build your first workflow in 5 minutes: HTTP Request → OpenAI → Slack
• Use AI nodes (OpenAI/Anthropic) to automate tasks like summarizing support tickets
• Schedule workflows with cron triggers (e.g., 0 9 * * * for daily 9 AM runs)
• Secure credentials with environment variables (.env file)

## 1. Self-Hosting with Docker

1.1 Install Docker

If you don’t have Docker installed, run:

# Linux/macOS
curl -fsSL https://get.docker.com | sh

# Windows (PowerShell)
winget install Docker.DockerDesktop

1.2 Run n8n

docker run -it --rm \
  --name n8n \
  -p 5678:5678 \
  -v ~/.n8n:/home/node/.n8n \
  n8nio/n8n

Expected output:

n8n ready on 0.0.0.0, port 5678
Version: 1.40.0
Editor is now accessible via: http://localhost:5678

Gotchas:

• Port conflict? Change -p 5678:5678 to -p 5679:5678 if port 5678 is in use.
• Permission errors? Add -e N8N_USER=node to run as non-root.

1.3 Persist Data

Add a volume to save workflows and credentials:

docker run -it --rm \
  --name n8n \
  -p 5678:5678 \
  -v ~/.n8n:/home/node/.n8n \
  n8nio/n8n

## 2. Your First Workflow

2.1 Access the Editor

Open http://localhost:5678 in your browser. You’ll see the visual workflow editor.

2.2 Create a "Hello World" Workflow

1. Click + New Workflow.
2. Drag an HTTP Request node from the left panel.
3. Configure it:
   • Method: GET
   • URL: https://api.github.com/users/n8n-io
4. Click Execute Node to test. You’ll see GitHub’s API response.

Expected output (truncated):

{
  "login": "n8n-io",
  "id": 45487711,
  "node_id": "MDEyOk9yZ2FuaXphdGlvbjQ1NDg3NzEx",
  ...
}

## 3. AI Agent Nodes (OpenAI/Anthropic)

3.1 Add an OpenAI Node

1. Drag an OpenAI node into your workflow.
2. Click Connect Account and enter your OpenAI API key (get one here).
3. Configure the node:
   • Resource: Chat
   • Operation: Create Chat Completion
   • Model: gpt-4o-mini
   • Prompt: Summarize this GitHub user data: {{ $json }}

3.2 Connect Nodes

1. Connect the HTTP Request node’s output to the OpenAI node’s input.
2. Execute the workflow. The OpenAI node will summarize the GitHub data.

Example output:

The GitHub user "n8n-io" is an organization account with ID 45487711. It was created to host the n8n project, an open-source workflow automation tool.

Gotchas:

• API limits? Use gpt-4o-mini for cost savings.
• Rate limits? Add a Wait node (e.g., 1 second delay) between requests.

## 4. HTTP Request and Webhook Nodes

4.1 HTTP Request Node

Use this to call any API. Example: Fetch weather data from OpenWeatherMap.

1. Add an HTTP Request node.
2. Configure:
   • Method: GET
   • URL: https://api.openweathermap.org/data/2.5/weather?q=London&appid=YOUR_API_KEY
   • Query Parameters:

     {
       "q": "London",
       "appid": "YOUR_API_KEY"
     }
     

     Copy

4.2 Webhook Node

Trigger workflows via HTTP. Example: Slack slash command.

1. Add a Webhook node.
2. Configure:
   • Path: slack-command
   • HTTP Method: POST
3. Click Webhook URLs to copy the URL (e.g., http://localhost:5678/webhook/slack-command).
4. Add a Slack node to send a response.

Test the webhook:

curl -X POST http://localhost:5678/webhook/slack-command \
  -H "Content-Type: application/json" \
  -d '{"text": "Hello from Slack!"}'

## 5. Scheduling and Cron Triggers

5.1 Add a Cron Trigger

1. Drag a Cron node into your workflow.
2. Configure:
   • Trigger On: Every Minute (or use cron syntax, e.g., 0 9 * * * for daily 9 AM).
3. Connect it to an HTTP Request node to fetch data daily.

Example cron expressions:

5.2 Manual Triggers

Use the Manual Trigger node to run workflows on demand.

## 6. Environment Variables and Credentials

6.1 Use .env File

Create a .env file:

echo "N8N_BASIC_AUTH_USER=admin" >> .env
echo "N8N_BASIC_AUTH_PASSWORD=securepassword" >> .env
echo "OPENAI_API_KEY=sk-your-key" >> .env

Restart n8n with the .env file:

docker run -it --rm \
  --name n8n \
  -p 5678:5678 \
  --env-file .env \
  n8nio/n8n

6.2 Secure Credentials

1. In the n8n editor, go to Credentials → Add Credential.
2. Select OpenAI and enter your API key.
3. Use the credential in any OpenAI node (no need to hardcode keys).

Gotchas:

• Never commit .env to Git! Add it to .gitignore.
• Rotate keys regularly (e.g., every 90 days).

## 7. Production Tips and Backups

7.1 Use PostgreSQL (Not SQLite)

For production, replace SQLite with PostgreSQL:

docker run -it --rm \
  --name n8n \
  -p 5678:5678 \
  -e DB_TYPE=postgresdb \
  -e DB_POSTGRESDB_DATABASE=n8n \
  -e DB_POSTGRESDB_HOST=postgres \
  -e DB_POSTGRESDB_PORT=5432 \
  -e DB_POSTGRESDB_USER=n8n \
  -e DB_POSTGRESDB_PASSWORD=securepassword \
  n8nio/n8n

7.2 Enable HTTPS

Use a reverse proxy like Nginx or Caddy:

# Caddy example
echo "localhost:5678 {
  reverse_proxy n8n:5678
}" > Caddyfile
docker run -d --name caddy -p 80:80 -p 443:443 -v $(pwd)/Caddyfile:/etc/caddy/Caddyfile caddy

7.3 Backup Workflows

Export workflows via the API:

curl http://localhost:5678/rest/workflows \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -o workflows_backup.json

Restore workflows:

curl -X POST http://localhost:5678/rest/workflows \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d @workflows_backup.json

7.4 Monitor Workflows

• Use the Execution List in the n8n UI to debug failed workflows.
• Enable Error Workflows to send Slack alerts on failures.

## Alternatives to n8n

## What's Next?

1. Build a real workflow: Automate a task like summarizing Slack messages with OpenAI.
2. Deploy to production: Use Kubernetes or n8n.cloud for scalability.
3. Explore AI features: Try n8n’s AI workflow generator (beta).

Need help scaling AI workflows? Hyperion Consulting specializes in AI tools and automation for enterprises. Visit hyperion-consulting.io to learn more.