Installing an application or a control panel on your Linux VPS

Published on 21/08/2026 Updated on 21/08/2026 3 views
VPS Linux Applications Docker

You have two routes: let the client area deploy a well-known application in one click, or install what you need yourself. The first saves time on a fresh server, the second leaves everything in your hands.

Which applications can be installed automatically on my VPS?

Some plans offer to install an application at delivery time. The catalogue includes Pterodactyl, Nextcloud, FastPanel, n8n, Coolify and Vaultwarden.

You pick the application at two moments:

  • when ordering, in the product form;
  • later, in the Réinstallation tab of your service — but remember that reinstalling erases everything already on the machine.

On the selection screen, a Rechercher une application… box helps you find one quickly, and an Aucune application card stays selected if you prefer a bare system. Applications incompatible with the chosen operating system appear dimmed.

What happens while my application is being installed?

If the application needs parameters, the client area first shows a "Configurez …" screen with the line "Fill in your application's parameters before installation.", the IP address and hostname of the VPS, then the application's own fields. The Lancer l'installation button starts the deployment.

The next screen announces "Installation en cours..." and notes that "this operation may take several minutes." The page reloads itself every thirty seconds; let it work.

Once finished, the service page shows an Application row among the VPS details, plus a card with the connection details specific to that application: address, administrator account, first-configuration notes. Write them down and change the suggested passwords immediately.

If it fails, the panel shows "Échec de l'installation" with an expandable Détails block. Copy that detail into a ticket: it contains the exact error.

How do I install an application that is not in the catalogue myself?

If what you need is not in the catalogue, install it yourself. On a fresh server, two approaches cover almost every case.

A classic stack, from distribution packages:

apt update
apt install nginx mariadb-server php-fpm php-mysql
systemctl enable --now nginx mariadb php8.2-fpm
mysql_secure_installation

Containers, easier to isolate, move and update:

curl -fsSL https://get.docker.com | sh
systemctl enable --now docker
docker compose version

Then put each application in its own compose.yaml, in its own directory. Six months later you will still know what is running and why.

What precautions should I take before putting an application online?

Four: expose only the ports that are genuinely public, put a domain name and HTTPS in front of any web interface, take a snapshot before a major upgrade, and back up the data outside the VPS.

  1. Publish only what must be published. Bind databases to 127.0.0.1 and open only the public service ports in the firewall. See the firewall and fail2ban guide.
  2. Put a domain name and HTTPS in front of every web interface. A Let's Encrypt certificate is one certbot command away and keeps your credentials off the wire in clear text.
  3. Take a snapshot before a major application upgrade: rolling back then takes minutes.
  4. Back up the data, not the system. A database exported regularly to storage outside the VPS will save you far more reliably than a full image you will never restore.

What do I do if my application will not start?

Five commands almost always reveal the cause: the state of the service, its log, the container logs, the ports already in use and the free disk space.

systemctl status SERVICE_NAME
journalctl -u SERVICE_NAME -n 100 --no-pager
docker compose logs --tail=100
ss -tulpn
df -h

Nine times out of ten the answer is in those five commands: stopped service, port already taken, full disk, or a configuration error spelled out in the log.

Was this article helpful?