A payments-ingestion service must keep running if an entire Google Cloud zone fails, must not lose acknowledged data, and must recover within 30 minutes. Which deployment best fits?
Show answer and explanation
Answer: B
A regional Dataflow job schedules its workers across multiple zones in the region, so losing one zone does not stop processing. Writing results to regional or multi-region storage keeps the data durable beyond any single zone. Failover happens without human action, which keeps recovery well inside the 30-minute target. Single-zone deployments fail along with their zone, and manual cut-over tends to take longer than the target while adding error risk.
Why the other options are wrong
- A. Pinning workers to one zone concentrates risk. If that zone fails, processing stops until it recovers, which breaks the requirement to survive the loss of a zone.
- C. A single-zone database with nightly backups can lose up to a day of data and needs restoration time after a zone failure, so it misses both the data and recovery targets.
- D. Manual traffic switching depends on an engineer acting quickly and correctly under pressure. It is not automatic and does not reliably meet the recovery target.
Exam tip. For zonal resilience, prefer regional or multi-region managed services over single-zone instances plus backups.