Server hardening for the VPS your app lives on.
A fresh VPS gets its first brute-force attempt within minutes of booting. Password SSH, no firewall, no backups and silent failed updates are how small servers die. We harden yours once, properly, then the platform watches it around the clock.
The boring, critical work most projects skip.
Everything below is standard practice and almost nobody does all of it. We do the full pass on Ubuntu, Debian and the usual panels (Hetzner, DigitalOcean, OVH, aaPanel, Ploi, Forge), and document every change.
- SSH: keys only, root login off, fail2ban with sane bans
- Firewall: default-deny, only your service ports open
- TLS: automatic issue and renew, HTTP redirected, HSTS on
- Backups: nightly, off-server, with a tested restore, not just a cron line
- Updates: unattended security patches with reboot windows
- Users and permissions: app runs unprivileged, sudo audited
Hardened and documented
Every change written down, nothing mysterious left behind.
What actually happens to unhardened servers.
Brute-forced SSH
Bots try thousands of passwords per hour against port 22 from the moment your server gets an IP. Password auth eventually loses; keys plus fail2ban ends the game.
Cryptominers and botnets
A compromised box mines crypto or joins a botnet quietly. You find out from a CPU alert, an abuse report from your host, or a blacklisted IP.
Backups that never restore
Half of all backup setups fail on first restore: wrong paths, expired credentials, silently full disks. We test the restore, because an untested backup is a hope, not a backup.
Expired certificates
A TLS certificate that lapses on a weekend logs your users out and floods you with browser-warning screenshots. Automatic renewal plus expiry monitoring makes it a non-event.
Unpatched CVEs
Last year's OpenSSH or kernel vulnerability stays exploitable forever on a server nobody updates. Unattended security patches close the window without you thinking about it.
No one watching
Every safeguard degrades silently: disks fill, services crash, bans stop firing. Monitoring is the difference between an incident and an outage.
App-level security too? See the security audit and the managed WAF.
One server or a fleet, hardened this week.
Tell us the OS, the host and what runs on it. Fixed price per server, documented handover, monitored after.