Exact VPS sizing, production Docker stacks & honest cloud TCO.
Model idle vs peak RAM for n8n, Coolify, Supabase, Immich, Ollama, and dedicated game servers. Measured AMD EPYC + Gen4 NVMe benchmarks with transparent Year-1 and renewal economics.

Baseline Stack: Hostinger KVM (−20%) & Proton Privacy ($20 USD Credit)
| Provider / Tier | Hardware & Architecture | List | Net Benefit | Code | Link |
|---|---|---|---|---|---|
Hostinger KVM KVM 2 | 2 vCPU AMD EPYC · 8 GB DDR5 · 100 GB NVMe · 8 TB | $8.79/mo | ~$7.03/mo (-20%) | Claim -20% | |
Proton (VPN & Privacy) VPN Plus / Unlimited | WireGuard + NAT-PMP (Gluetun Docker) · Swiss No-Logs · Mail/Drive/Pass | Subscription | $20 USD credit | Claim $20 |
5 Interactive Sizing Instruments

n8n VPS Sizer
Calculate exact RAM, vCPU, NVMe storage, and Postgres/Redis Queue Mode thresholds based on active workflows, daily executions, and webhook concurrency.

Docker Stack Sizer
Combine Coolify, Supabase, Nextcloud, Immich, WordPress/WooCommerce, Odoo, n8n, and Gluetun VPN to compute exact idle vs peak RAM and auto-generate Docker memory limits.

Game Server Sizer
Compute RAM, single-core CPU clock requirements, NVMe world-save throughput, and monthly egress for Minecraft Paper/Forge/Fabric, Palworld, Rust, CS2, Valheim & Terraria.

LLM VRAM & GPU
Calculate exact model weights + KV-cache memory for Ollama, OpenWebUI, and vLLM (Llama 3.3, DeepSeek-R1, Qwen 2.5) and decide between a KVM VPS or hourly Cloud GPU.

Cloud TCO & Savings
Compare 12–36 month real Total Cost of Ownership including AWS egress fees ($0.09/GB), backup surcharges (+20%), renewal rates, and our -20% Hostinger KVM referral stacking.
Docker Compose Blueprints & Real Hardware Costs

How to Self-Host n8n on a VPS in 2026: RAM, Docker Compose & Real Cost
Self-hosting n8n Community Edition with PostgreSQL requires 2 GB RAM minimum and 4–8 GB RAM for reliable production webhooks. On Hostinger KVM 1 (4 GB) or KVM 2 (8 GB) with referral code XF3PLUTARXVF (-20%), total cost drops to ~$5.19–$7.03/mo with unlimited executions — saving $180–$500+/yr vs n8n Cloud.

Self-Hosting Coolify on a VPS (2026): Replace Vercel & Heroku for $7/mo
Coolify turns any Ubuntu 24.04 KVM VPS into a self-hosted Vercel/Railway alternative with push-to-deploy, preview PRs, and managed Postgres/Redis. Start on KVM 2 (8 GB RAM, ~$7.03/mo with code XF3PLUTARXVF) to host 5–10 Next.js/Node apps plus databases comfortably.

Gluetun + ProtonVPN WireGuard on a VPS (2026): Docker Kill Switch & NAT-PMP
Gluetun consumes ~45 MB RAM and wraps any Docker container behind an encrypted ProtonVPN WireGuard tunnel with a kernel-level kill switch (`network_mode: service:gluetun`), ensuring your real VPS IP never leaks. New Proton subscribers also receive US$20 in account credits with referral code 5AA9X0DA (pr.tn/ref/5AA9X0DA).

Minecraft Paper Server on a VPS (2026): RAM, Aikar Flags & Real Cost
A Minecraft 1.21+ Paper/Purpur server with 15–25 players and 20 plugins needs 6 GB JVM heap (`-Xmx6G`) inside an 8 GB VPS so Linux & Docker overhead never triggers OOM kills. Hostinger KVM 2 (8 GB DDR5, NVMe, Wanguard DDoS) drops to ~$7.03/mo with code XF3PLUTARXVF (-20%).

Self-Hosting Coolify + n8n on One VPS (2026): Memory Budget & Setup

Self-Hosting OpenWebUI + Ollama (2026): CPU KVM VPS vs Cloud GPU Cost

Self-Hosting Supabase on a VPS (2026): Real RAM, Docker Stack & Cost

Self-Hosting Immich on a VPS (2026): RAM for Facial ML, NVMe & Real Cost

Palworld Dedicated Server on a VPS (2026): Memory Leak Guard, RAM & Cost

Fix Cloudflare Error 522 Connection Timed Out on Production VPS Nodes (2026)
Real Monthly Cost for 8 GB RAM

| Provider | Backups | Yr 1 | Renew |
|---|---|---|---|
| Hostinger KVM | Free weekly automated snapshots included | $7.03 | $12.99 |
| Hetzner Cloud | +20% surcharge for automated backups | $8.50 | $8.50 |
| Vultr High Frequency | +20% surcharge for auto backups | $48.00 | $48.00 |
| DigitalOcean | +20% surcharge for weekly backups | $56.00 | $56.00 |
| AWS EC2 (t3.large + EBS) | EBS Snapshots billed per GB/month | $78.40 | $78.40 |
Reproducible hardware telemetry built for engineers & AI search.
Every guide ships with measured idle vs peak RAM, memory-bounded docker-compose.yml files, transparent renewal pricing, and open machine-readable indices at /llms.txt and /llms-full.txt.