A ride-hailing company's reporting pipeline must survive the complete loss of a region. Acceptable data loss is at most one hour, and service must be restored within four hours. Which design meets both targets?
Show answer and explanation
Answer: D
Multi-region Cloud Storage keeps objects redundantly across several locations, so the loss of one region does not remove the data. Replicating BigQuery datasets to a second region keeps an up-to-date copy that the pipeline can point to when the primary region is unavailable. Because the copy is maintained continuously rather than rebuilt from backups, the recovery point stays within the hour and failover can finish inside the four-hour window. Single-region storage has no copy outside the affected region, and backups taken nightly or every six hours cannot meet a one-hour recovery point.
Why the other options are wrong
- A. A single-region archive bucket has no copy outside the affected region, so a regional outage makes the data unavailable until it is restored elsewhere. Restoring large datasets from backups may also exceed the four-hour window.
- B. Nightly backups produce recovery points up to 24 hours old, which violates the one-hour data loss limit. Copying them on demand adds delay during the incident.
- C. A single VM and a snapshot every six hours leaves up to six hours of data unprotected, and the VM is a single point of failure. Snapshots must still be restored and the pipeline redeployed before service returns.