COF-C03 Question 196
Single answerData lineageA data engineering team loads a SALES_RAW table each night and then creates a SALES_CURATED table with SQL transformations. A BI team later creates a secure view named VW_MONTHLY_SALES on top of SALES_CURATED for reporting. During an audit, the company must quickly determine which downstream objects are affected if the SALES_RAW table structure changes. Which Snowflake feature should the team use to trace these upstream and downstream object relationships most directly?
- A
Use Snowsight Lineage to visualize dependencies between the base table, derived table, and secure view
- B
Use Time Travel to inspect historical versions of the SALES_RAW table and infer dependent objects
- C
Use a zero-copy clone of the database and compare cloned objects to identify lineage relationships
- D
Use the load history for SALES_RAW to identify all downstream views and tables built from it
Show answer and explanation
Correct answer: A
Explanation
The best answer is to use Snowflake Lineage in Snowsight. Data lineage is designed to show how data moves through Snowflake objects and how upstream and downstream objects are connected, which supports impact analysis, governance, and auditing. In this scenario, the team needs to identify dependencies from SALES_RAW to transformed tables and reporting views. Time Travel, cloning, and load history are all useful Snowflake capabilities, but none of them directly provides dependency visualization for lineage analysis. This aligns with Snowflake documentation and best practices around using Snowsight governance features to understand object relationships and assess the effects of schema or pipeline changes.
- A. Correct.
Correct. Snowflake provides Lineage in Snowsight to help users trace object relationships across upstream and downstream dependencies. In this scenario, it is the most direct way to see how SALES_RAW feeds SALES_CURATED and then VW_MONTHLY_SALES, which is exactly what is needed for impact analysis during an audit or schema change review.
- B. Incorrect.
Incorrect. Time Travel lets users access historical data from tables, schemas, and databases within the retention period, but it does not provide dependency mapping between objects. It helps with recovering or querying past data states, not with tracing lineage across derived objects.
- C. Incorrect.
Incorrect. Zero-copy cloning creates a copy of existing objects without duplicating underlying storage initially, but it is not intended to discover lineage. Comparing cloned objects may help with testing changes safely, yet it does not reveal upstream and downstream dependency relationships.
- D. Incorrect.
Incorrect. Load history is useful for understanding data loading activity, such as COPY INTO operations and file load details, but it does not show which downstream tables or views were created from a source table through SQL transformations. It is an operational history feature, not a lineage feature.