You are managing a production Delta Lake table, sales_data, which is used by multiple downstream analytics jobs. The table is partitioned by region and sale_date. Recently, your team needs to ingest new data files that include two additional columns: discount_code (string) and promotion_flag (boolean), which were not present in the original schema.

Requirements:

Zero Downtime: The table must remain available for reads and writes during the schema update.
Data Consistency: All new data ingested must include the new columns, and existing queries should not break.
Auditability: You must be able to track when and how the schema was changed.
Best Practices: The solution should follow Delta Lake best practices for schema evolution and production data management.

Which approach best satisfies all requirements? Select the best answer and explain why the other options are less suitable.