The problem
Our backups were well built for fire and flood, but they were never designed around an attacker who deletes the backups first. Nobody had proven that the business systems could be brought back, or written down how.
What I did
- Ranked the systems the business can’t run without, starting with the ERP and shipping: if we can’t ship, it doesn’t matter what we produce.
- Built a recovery environment that can restore core systems without the production network, and kept it current with an automatic weekly restore of the core databases.
- Ran restore drills that passed for the ERP system, the other SQL Server databases and the in-house web apps, including a fully offline run.
- Wrote printed runbooks for general disaster recovery and for a ransomware attack, including a recovery path that needs no scripts or AI to follow. I collected them with the response plan in a single emergency binder.
- Set up an offsite recovery kit that refreshes automatically, backed by immutable offsite copies.
- Documented break-glass access so recovery doesn’t depend on any one sign-in system being available.
Outcome
Recovery is tested instead of assumed, with an automatic weekly restore and hands-on drills planned through the year.