◆
Rehearsed, not written
Recovery is something you have done, repeatedly, rather than something you have described.
Most disaster recovery is a document. It describes a procedure nobody has performed, with assumptions nobody has tested, and it will be read for the first time on the worst day of the year.
Used for recovery, continuity, backup verification and failover.
What this looks like in practice
Recovery is something you have done, repeatedly, rather than something you have described.
A backup nobody has restored is a hypothesis. Verification happens without anyone scheduling it.
How much work a failure could cost you, measured in moments rather than in hours of re-entry.
Proof that recovery works, drawn from exercises that actually happened.
Every organisation that has had a bad recovery had a recovery plan. The plan was not the problem; the fact that it had never been executed was.
The goal of a recovery exercise is for it to be dull. Everyone knows the steps because they have done them, the recovery point is the one documented, and the exercise ends early. A dramatic drill is not a successful one — it is a preview.
Backups verified by being restored, which is the only verification that means anything.
Recovering to a moment rather than to last night’s snapshot.
Regional and provider failover exercised before the day it matters.
Continuity proof that comes from exercises instead of from a drafting session.
We had a real incident eight months in. It went exactly like the drill, including how long it took. Nobody had to make a decision under pressure.
Moments rather than hours for most workloads — and more importantly, a number you have measured during an exercise rather than one in a document.
Because they are restored. A backup that has never been restored is an assumption, however well-monitored the backup job is.
Often enough that it is unremarkable. Teams that exercise quarterly handle real incidents noticeably better than teams that exercise annually.
The evidence is a record of exercises you actually ran, which is stronger than a plan. Where a specific attestation is required, ask us directly.
Start on something that will still be right at a hundred times your size, and pay now what a startup should pay.
Move in stages with the old arrangement still running, and cut over only once the replacement has carried real traffic.
Get capacity when the experiment is ready, keep what produced a result, and put models in front of users without a separate platform.
Keep everything at full fidelity, query it without a project, and stop deciding today what you will want to ask next year.
Ship the application your users touch, on an address that stays put, with every version one click behind you.
Build the tools your own teams need, at a cost that makes small improvements worth doing.
Begin with one workload. Keep what you have running. Decide whether to continue once you have seen it carry something real.
No migration required to begin. Nothing to tear down if you stop.