You found n8n, set up a few workflows, and everything worked locally. Then you moved it to a cheap VPS, and now your automation dies every time a webhook fires. Sound familiar?
A cheap VPS isn’t the problem. Picking the wrong one is. n8n is lightweight, but it needs specific resources to run reliably. This checklist helps you choose a cheap VPS that actually works for n8n, without guessing.
Why a cheap VPS for n8n is a smart move (if you pick right)
Running n8n on your laptop means your automation stops when you sleep or when your ISP has a hiccup. A cheap VPS solves that. It keeps your workflows running 24/7, handles webhooks, and connects APIs without your PC being on.
But cheap VPS providers often oversell resources. A $3 plan may look good on paper but can’t handle n8n’s Node.js process and a database at the same time. You need to know what to check before you buy.
Step-by-step checklist for choosing a cheap VPS for n8n
1. Confirm enough RAM for n8n + your database
n8n itself uses around 256–512 MB of RAM when idle. If you store workflow data in SQLite (n8n’s default), you need at least 1 GB of RAM total. If you plan to use PostgreSQL or MySQL on the same instance, aim for 2 GB.
- 1 GB RAM: Good for 10–20 simple workflows, no external database.
- 2 GB RAM: Handles 30–50 workflows with SQLite or a small PostgreSQL.
- 4 GB RAM: Overkill for most, but fine if you run heavy integrations.
2. Check the virtualization type (KVM only)
Shared hosting or OpenVZ containers often throttle CPU and memory under load. n8n runs as a Node.js process, which can spike CPU during workflow execution. You need KVM virtualization for guaranteed resources.
- Avoid: OpenVZ, LXC, shared hosting.
- Look for: KVM, KVM-based, or bare metal.
3. Verify you can install Docker or Node.js manually
Some cheap VPS providers lock you into a control panel with pre-installed apps. n8n needs either a manual Node.js install or Docker. Make sure you have root SSH access and can run apt, yum, or docker commands.
- Must have: Full root access, no forced app store.
- Nice to have: Pre-installed Docker (saves time).
4. Look for SSD storage with decent I/O
n8n stores workflow data, logs, and sometimes file-based queues on disk. If the VPS uses old HDDs or throttled SSDs, your workflows will slow down or freeze.
- Minimum: 20 GB SSD.
- Check: Can you access disk I/O benchmarks? Look for 100+ MB/s sequential read.
5. Test with a short billing cycle first
Don’t commit to a year upfront. Many cheap VPS providers have monthly plans under $10. Test n8n for a week. If it crashes, you lose only $5, not $60.
- Preferred: Monthly billing.
- Avoid: Prepaid 12-month plans from unknown hosts.
Three common mistakes that break n8n on a low-cost VPS
Mistake 1: Choosing a VPS with less than 1 GB RAM
You see a $3 plan with 512 MB RAM and think “n8n is small.” Then you start a webhook workflow that calls three APIs simultaneously. The process hits the swap file hard, and n8n becomes unresponsive.
Mistake 2: Forgetting about swap space
Even with 1 GB RAM, n8n can spike during heavy workflow execution. If your VPS has no swap configured, the process gets killed by the OOM (out-of-memory) killer. Always add at least 1 GB of swap after installation.
Mistake 3: Ignoring CPU limits
Some cheap VPS providers advertise “unlimited CPU” but throttle after a few seconds of sustained use. n8n workflows that loop or process large data sets will trigger throttling. Look for a host that publishes CPU benchmarks or offers dedicated vCPU cores.
Mini scenario: How a $7 VPS handled 50 workflows with zero downtime
A beginner needed to run 50 n8n workflows that connected their CRM, email, and Slack. They bought a $7/month KVM VPS with 1 GB RAM, 25 GB SSD, and a single dedicated vCPU.
- Installed n8n via Docker.
- Added 1 GB swap.
- Configured n8n to use SQLite.
After 90 days of uptime, the VPS handled 68,000 workflow executions with zero crashes. The only downtime was a planned reboot for a kernel update. The $7 plan was enough because the workflows were simple (one trigger, one or two actions each).
For heavy workflows with file processing or database joins, the same user would need a 2 GB plan. But for typical API automations, $7/month was plenty.
