Disaster Recovery Drill

Recovery planning is usually described with two numbers. RPO is how much data can be afforded to lose, measured as time since the last good backup. RTO is how long the business can afford to be down before service is restored. Those two numbers sound simple, but they only mean something once each one ties to something observable. RPO ties to which recovery point gets selected. RTO ties to how much elapsed time passes between starting a restore and proving it actually worked.

This drill puts both numbers under pressure at once, alongside a budget constraint, in a scenario where the easy path, just restore the latest backup, doesn't actually satisfy what's being guaranteed.

Here's the scenario. RPO is 15 minutes, RTO is 2 hours, and the budget is tight. But the last good backup is unclear, because recent jobs show mixed status. Under those constraints, a usable plan chooses a specific recovery point, restores into a specific target, and proves progress with signals that can be observed while the restore is running.

Success in this drill is an ordered runbook, not a paragraph. Each step must state what changes in the environment, and what ...