Explains recovery objectives and impact tolerances
RPO and RTO per data set must meet the impact tolerance of every service depending on it (DORA, ISO 22301).
Can AI do this?
AI does this partly Some of the tested AI models master this; others do not show it. 2 of 4 models in the current benchmark master it.
| Model | Tier | Probability of mastery | Status |
|---|---|---|---|
| gemini-4-argon | Flagship | 100% | Mastered |
| gemini-3.5-flash-lite | Low cost | 100% | Not shown |
| gpt-6-luna | Low cost | 100% | Not shown |
| claude-haiku-5-5 | Low cost | 100% | Mastered |
Mastered means a probability of at least 90% that the model has this knobit, estimated from its graded answers. “Not shown” means the answers do not settle it either way. How the models are tested.
What it looks like in practice
Expected result: RPO and RTO per data set must meet the impact tolerance of every service depending on it (DORA, ISO 22301).
How it is assessed: Answers a written question on the recovery objectives correctly
Related knobits
Needed for
- O-301 Recognises an RPO the copies cannot deliver
- O-302 Recognises an RTO longer than a service's impact tolerance
- O-307 Recognises a critical data set with no recovery objectives
- O-308 Recognises a critical service depending on a data set outside the plan
- O-310 Recognises a restore slower than the RTO
- O-331 Reconciles recovery objectives with copies and tolerances
- O-353 Explains restore testing
All 14 knobits of Plan backup, recovery and continuity.
Source: Practice the stage 3 reference and review checklist rest on