Disaster Recovery (DR)

Disaster recovery is usually described with two numbers: RPO (Recovery Point Objective — how much data can be lost, measured as time since the last good backup) and RTO (Recovery Time Objective — how long systems can be down before the business is meaningfully harmed). A DR plan built around a 24-hour RPO and a 4-hour RTO implies very different backup frequency and infrastructure than one built around near-zero RPO and RTO.

DR is broader than backup, even though the two get conflated. A backup is a copy of data; a DR plan is what happens with that copy when the primary system is gone entirely — where does it restore to, who executes the plan, in what order do systems come back online, and has the whole sequence actually been tested end to end rather than just assumed to work.

Ransomware changed DR planning in one specific way: recovery has to assume the backup itself might be a target. Immutable backups (that can't be altered or deleted, even by an account with admin credentials, for a defined retention window) and backups stored outside the primary network became standard practice specifically because attackers started encrypting or deleting backups as part of the attack, not just the primary systems.