What is the expected behavior when an upstream system emitting change data logs encounters a missing column in the target schema?

When an upstream system processes change data logs and encounters a missing column in the target schema, the expected behavior is that the new nested field will be read as NULL for existing unmatched records. This happens because, during the ingestion or processing of change data, if a specific column expected in the target schema is not present in the incoming data, the system cannot populate that column for records that do not have corresponding data.

This is a common approach in data engineering where schemas can evolve over time. When records lack certain fields, it is standard practice for databases and data lakes to assign NULL to those fields for any existing data that does not match, ensuring the structure of the dataset remains consistent. Consequently, this approach allows the target system to maintain compatibility with future data that might include the previously missing column, all while keeping existing records intact.

The other scenarios provided in the choices do not accurately reflect standard behavior when handling missing columns in target schemas aligned with data engineering practices. For instance, throwing an error due to unsupported changes would not permit the data to be ingested and processed effectively. Storing updates in a "rescued" column might complicate the data structure without serving an immediate purpose. Automatically filling missing fields with default values could lead to misleading interpretations of