Data scientists running reports against a Cloud SQL for PostgreSQL orders database are impacting website performance. Provide up-to-date order datasets to analysts without affecting the site. What should you do?

Using Datastream to replicate Cloud SQL data into BigQuery provides a real-time, fully managed change data capture (CDC) solution. BigQuery is optimized for analytical queries, allowing data scientists to run complex reports without impacting the operational Cloud SQL database. This approach ensures up-to-date data for analysis and maintains website performance. Exporting to CSV via Cloud Functions is less efficient for continuous, near real-time updates and requires manual management of data loading into an analytical store. Dual-writing to Bigtable is not ideal because Bigtable is a NoSQL wide-column store, not optimized for complex analytical queries typically performed by data scientists. Cloning the PostgreSQL instance every 15 minutes is resource-intensive, incurs high costs, and still requires analysts to query an operational database, potentially impacting performance or requiring further data movement.