An internal analytics workload runs in a single Region and is not customer-facing. The business accepts an RTO of up to twenty-four hours and an RPO of up to one day after a regional outage, and the overriding constraint is to spend almost nothing on standby infrastructure between incidents. The team already copies nightly application backups and database snapshots to a second Region. After a regional outage they are willing to provision the entire environment from those copies before the workload returns to service. Which disaster recovery strategy best matches these requirements?
- AWarm standby, running a permanently on but scaled-down copy of the whole environment in the second Region so it can absorb production load shortly after failover.
- BBackup and restore, copying backups to the recovery Region and provisioning the whole environment from those copies only after a disaster is declared. Correct
- CPilot light, keeping the database continuously replicated to a live standby instance in the recovery Region while the application tier remains switched off until failover.
- DMulti-site active-active, serving the analytics workload from both Regions at the same time with traffic routing that fails over automatically when one Region degrades.
Why A is wrong: Warm standby keeps an always-running scaled-down stack, which adds continuous cost that conflicts with spending almost nothing on standby, and its fast RTO exceeds what these relaxed objectives require.
Why B is correct: Backup and restore provisions nothing until failover and relies on cross-Region backup copies, giving the lowest standing cost while meeting a relaxed RTO of hours and an RPO measured in a day.
Why C is wrong: Pilot light needs a continuously replicated, always-on data layer, which costs more than nightly backup copies and delivers a tighter RPO than the one-day objective actually demands.
Why D is wrong: Active-active runs full duplicate capacity serving live traffic, the most expensive option, which directly contradicts the requirement to minimise standby spend for a tolerant internal workload.