Scenario: You are running a Dataflow streaming job to read messages from a message bus that does not support exactly-once delivery. After applying transformations, the results are loaded into BigQuery. You want to make sure that data is streamed into BigQuery with exactly-once delivery semantics. You expect an ingestion throughput of about 1.5 GB per second into BigQuery. Question: What should you do to ensure that data is streamed into BigQuery with exactly-once delivery semantics? **Not part of question** Quota limit might change with time. You do not need to worry about the quota limit numbers. As of Nov 2024, the throughput quota has been updated to - 3 GB per second throughput in multi-regions; 300 MB per second in regions