Why I run n8n on a VPS and regret absolutely nothing
Turns out owning your automation stack has vibes. Also control. Mostly vibes though.
Everyone asks why I don’t just use Make or Zapier.
The answer is embarrassingly simple: I wanted to own the thing.
The actual setup
Hostinger KVM2 VPS, Ubuntu, Docker Compose. n8n runs on port 5678 behind nginx. SSL via certbot. Takes maybe 20 minutes to set up if you’ve done it once.
services:
n8n:
image: n8nio/n8n
restart: unless-stopped
ports:
- "5678:5678"
environment:
- N8N_HOST=n8n.yourdomain.com
- N8N_PROTOCOL=https
- WEBHOOK_URL=https://n8n.yourdomain.com
volumes:
- n8n_data:/home/node/.n8nThat’s most of it. The rest is nginx config and certbot commands you can find anywhere.
Why self-hosted wins for my workflow
Cost: At my usage level, self-hosted is $6/month (VPS). Cloud n8n is $20-50+. Math works.
No execution limits: I have some workflows that run every minute. That would be expensive in the cloud. It’s just CPU on my VPS.
Data stays mine: Some of my workflows touch client data. I don’t want it going through a third-party platform I don’t control.
I can break it: This sounds bad but it’s actually good. When something breaks, I understand the stack well enough to fix it. Black box SaaS failures are much harder to debug.
The real downside
Maintenance is mine. When n8n releases a security update, I have to run docker compose pull && docker compose up -d. This happens a few times a year. It takes 2 minutes.
I genuinely don’t understand why this is treated as a dealbreaker.
If you’re running more than 20 workflows and spending real money on a cloud automation platform, the math almost certainly favors a VPS. Run the numbers.