Deploy Next.js on a VPS (PM2 + Nginx)

Next.js on a VPS vs Vercel (Cost Math)

Four stages, one evening:

1. Prepare Ubuntu — update packages, create a deploy user with sudo, install Node LTS via NodeSource, open 22/80/443 in UFW.
2. Build and run with PM2 — npm ci && npm run build, then pm2 start npm --name myapp -- start; save the process list for reboots.
3. Nginx reverse proxy — server block forwarding port 80 to your Node port with proxy headers set.
4. Domain and SSL — point DNS, then certbot --nginx -d yourdomain.com for free Let's Encrypt HTTPS with auto-renewal.

The economics versus Vercel: a ₹5,900/month cloud VPS carries unmetered-ish bandwidth and any number of apps, while Vercel's metered egress grows with traffic. Below production scale, the VPS wins cost; above it, you own capacity either way.

Step 1: Prepare Ubuntu VPS (Node, UFW)

How this plays out with Next.js: settings that look cosmetic usually decide page speed and uptime under real traffic. The defaults WebJi® ships are tuned for Indian network conditions, and support can review your setup on request.

Step 2: Build and Run With PM2

Build and Run With PM2. Do this on one site or a staging copy first, then check the public URL in a private window so cache does not hide the result. Stop if the page errors, and email support@webji.services with the step name and the exact error before you repeat it on every account. WebJi® support mail is staffed 24/7 in Hindi and English; the phone line +91 141 4941811 is business hours only.

Step 3: Nginx Reverse Proxy Config

How this plays out with PM2 process manager: settings that look cosmetic usually decide page speed and uptime under real traffic. The defaults WebJi® ships are tuned for Indian network conditions, and support can review your setup on request.

Step 4: Domain and SSL With Certbot

Domain and SSL With Certbot. Install the certificate on the account that actually serves the site, force HTTPS, then view the page source for http:// images, stylesheets or scripts. The padlock stays broken until those are HTTPS too. Renewal should be automatic; if the padlock disappears, check that the certificate is still bound to this hostname.

Keeping It Alive: Logs, Restarts, Updates

Production discipline is four habits:

1. Logs — pm2 logs for app output, nginx access/error logs for the front; logrotate or journald limits so disks do not fill silently.
2. Restarts that survive reboots — pm2 save plus pm2 startup registers the process list with systemd; without it, a server reboot leaves your app dark until you notice.
3. Updates — deploy via Git pull and rebuild in a release folder, flipping a symlink atomically; roll back by flipping back.
4. Backups and monitoring — snapshots before deploys, and uptime monitoring pinging the homepage so you learn about outages before users tweet them.

The failure modes are predictable — full disks, dead PM2, expired certificates — and all three are covered by habits, not heroics. Managed cloud, explained, covers handing the boring parts over.