Your target is an RTO of 5 minutes and an RPO near zero for a critical database. Which approach fits?
Why C is the answer
A synchronous, multi-zone or multi-region database with automatic failover
A near-zero RPO needs synchronous replication, and a five-minute RTO needs automatic failover.
In detail
- Near-zero data loss needs synchronous replication.
- Automatic failover meets the 5-minute target.
- Synchronous replication adds write latency.
Use case
An illustrative scenario, not a real incident.
A payments database writes synchronously to another zone, so a failure loses almost nothing, and failover is automatic.
See it in different ways
flowchart LR A["Backup and restore: cheapest, slowest"] --> B["Warm standby"] B --> C["Active-active: costliest, fastest"]
stateDiagram-v2 [*] --> Targets Targets --> BackupRestore: hours are acceptable Targets --> WarmStandby: minutes Targets --> ActiveMultiRegion: near zero loss ActiveMultiRegion --> AutomaticFailover AutomaticFailover --> [*]
mindmap
root(("Recovery targets"))
n1["RPO near zero"]
n2["Synchronous replication"]
n3["RTO of 5 minutes"]
n4["Automatic failover"]Key point: Near-zero RPO needs synchronous replication.
Why the other options are wrong
- A. Rebuilding the database from infrastructure as code after a failure
- Rebuilding takes too long.
- B. Nightly backups copied to another region for safekeeping
- Nightly backups lose up to a day.
- D. Weekly exports to Cloud Storage with a tested restore process
- Weekly exports lose far more.