The problem
Ransomware protection depended on someone swapping external drives every working day. A missed swap could fill a drive, and the only way to check backup health was to sign in to the backup server.
What I did
- Added immutable offsite copies, so backups can’t be deleted or encrypted, even by an attacker with admin rights.
- Cut the drive routine from a daily swap to a weekly offsite rotation, with scripts that eject and remount drives automatically.
- Published a backup status feed every 30 minutes that can be read without credentials or a server login.
- Diagnosed a Hyper-V integration bug that broke application-consistent backups on new Windows Server 2025 VMs, and applied a workaround until the vendor fix shipped.
- Built an isolated recovery environment and ran restore drills against it.
Outcome
Backups are protected offsite, need less manual handling, and problems show up within half an hour instead of whenever someone checks.