A team sets very strict data-loss tolerance but ignores intersite commit latency. What should guide protection-mode selection?

1. Assume asynchronous transport always acknowledges remote durable storage before commit.
2. Select solely from the number of PDBs.
3. Choose the longest region name to guarantee resilience.
4. Evaluate RPO, transport mode, network latency, and availability tradeoffs together.