n8n Production Checklist: 9 Things to Set Up Before Going Live

You've built your workflows, tested them a few times, and they work. Before you put them in front of real data and real clients, there are a few things worth setting up that will save you from 3 AM wake-up calls later.

The checklist

1. Switch to PostgreSQL. SQLite works for development, but it doesn't handle concurrent executions well. If you're running more than a handful of workflows on a schedule, PostgreSQL is the safer bet for execution data storage.

2. Set execution data pruning. n8n stores every execution by default. Without pruning, your database will grow until it slows down or fills the disk. Set EXECUTIONS_DATA_MAX_AGE to something reasonable — 168 hours (7 days) is a good start.

3. Enable queue mode for heavy loads. If you're running 50+ workflows, queue mode with Redis separates the main process from execution workers. This prevents a single heavy workflow from blocking everything else.

4. Secure your instance. Set N8N_BASIC_AUTH_ACTIVE=true at minimum. Better: put n8n behind a reverse proxy with HTTPS and authentication. An exposed n8n editor is a security incident waiting to happen.

5. Back up your workflows. Export workflows to JSON and store them in version control. If your server dies, you want to be able to restore everything in minutes, not hours.

6. Set up error handlers. Add an Error Trigger workflow that routes failures to email or Slack. Assign it to every production workflow. This catches execution-level errors and gives you a first line of defense.

7. Add external monitoring. Error handlers catch errors — but they don't catch silence. RootBrief monitors your workflows from outside n8n. If a workflow stops running entirely, or n8n itself goes down, you still get alerted. Starting free — with the Pro plan at $19/mo for Slack integration and AI monitoring (1,000 events/mo), and Pro Max at $100/mo for teams — this is one of the cheapest insurance policies you can buy for production automation.

8. Test with production-like data. A workflow that works with 10 rows might time out with 10,000. Test with realistic volumes before going live.

9. Document your workflows. Label your nodes clearly. Add sticky notes explaining what each section does. Future you — or the person who inherits these workflows — will thank you.

Step 7 takes two minutes. RootBrief connects to your n8n API key and starts monitoring all workflows immediately. start free.