In this article
Every server exposed to the internet faces automated scans and password-guessing attacks from the moment it goes live. The steps below block most of them from the start.
Restrict access
- Use SSH keys instead of passwords and disable direct root login.
- Where possible, open admin panels only to specific IP addresses.
- Use a tool that blocks repeated failed logins (such as Fail2ban).
Firewall
Keep open only the ports you actually use: 80 and 443 for the web, SSH for administration. The database port (3306) should not be open to the internet; the application should connect to the database from inside the server.
Keep it updated
Apply security updates for the operating system, web server, PHP and database regularly. Most attacks target holes that were patched months earlier.
Least privilege
- Give each application its own database user with access only to its own database.
- Do not grant wider file permissions than needed, and disable script execution in upload folders.
- Keep secrets (database passwords, API keys) outside the web root.
Monitoring and backups
Watch disk, CPU and RAM usage, error logs and failed login attempts. Regular backups stored off the server are your last line of defence when everything else fails.
Frequently asked questions
Does ADK apply these measures on the servers it sets up?
Yes. A firewall, key-based access, updates and separate database users are standard on the cloud servers we set up for our software plans.
How can I tell if my server is under attack?
Sudden CPU or traffic spikes, processes you do not recognise, changed files and floods of failed logins in the logs are the main signs.
This article is for general information. Contact our support team for actions specific to your account.

