ADA-C01 Question 420
Single answerMonitor and assess automatic clustering usageA Snowflake administrator enabled Automatic Clustering on several large fact tables that use clustering keys. After a week, finance reports a noticeable increase in compute charges. The administrator needs to determine which tables are driving Automatic Clustering activity and assess whether the clustering effort is justified. Which approach will provide the most accurate account-level view of Automatic Clustering usage for this analysis?
- A
Query the ACCOUNT_USAGE.AUTOMATIC_CLUSTERING_HISTORY view to review credits consumed and rows/bytes reclustered by table over time.
- B
Query the WAREHOUSE_METERING_HISTORY view and filter for warehouses used by ETL jobs, because Automatic Clustering runs on user warehouses.
- C
Use SHOW TABLES to review clustering keys and table sizes, because tables with clustering keys automatically indicate the exact credits consumed by Automatic Clustering.
- D
Query QUERY_HISTORY for ALTER TABLE ... CLUSTER BY statements, because Automatic Clustering only runs when manually triggered by DDL changes.
Show answer and explanation
Correct answer: A
Explanation
To monitor and assess Automatic Clustering usage, administrators should use Snowflake Account Usage views designed for serverless features, especially ACCOUNT_USAGE.AUTOMATIC_CLUSTERING_HISTORY. This view helps determine which tables are incurring reclustering work and how much credit consumption is associated with that activity over time. That makes it the best starting point when evaluating whether clustering keys are delivering enough query-pruning benefit to justify their maintenance cost. By contrast, warehouse metering views track customer-managed warehouses, not Snowflake-managed serverless services such as Automatic Clustering. Metadata commands like SHOW TABLES can reveal whether clustering keys exist, but they do not quantify usage or cost. This aligns with Snowflake documentation and best practices for monitoring serverless feature consumption through Account Usage views before deciding whether to keep, modify, or remove clustering keys.
- A. Correct.
Correct. ACCOUNT_USAGE.AUTOMATIC_CLUSTERING_HISTORY is the appropriate source to monitor and assess Automatic Clustering usage at the account level. It allows administrators to analyze activity over time and attribute reclustering work and associated credit consumption to specific tables. This is the most direct way to identify which tables are generating Automatic Clustering cost and whether that maintenance activity appears significant relative to expected performance benefits.
- B. Incorrect.
Incorrect. This reflects a common misconception. Automatic Clustering does not consume credits from a customer-managed virtual warehouse. Snowflake performs this as a serverless service, so warehouse metering is not the authoritative source for Automatic Clustering usage. WAREHOUSE_METERING_HISTORY is useful for warehouse workloads, but not for isolating serverless Automatic Clustering charges by table.
- C. Incorrect.
Incorrect. SHOW TABLES can help identify whether tables have clustering keys and provide some metadata, but it does not report Automatic Clustering consumption or historical reclustering activity. Knowing a table has a clustering key does not reveal whether Snowflake has spent significant credits maintaining it.
- D. Incorrect.
Incorrect. Automatic Clustering is not dependent on repeated manual ALTER TABLE ... CLUSTER BY execution. After a clustering key is defined and Automatic Clustering is enabled, Snowflake maintains clustering automatically in the background as data changes. QUERY_HISTORY of DDL statements would not provide a reliable measure of ongoing Automatic Clustering usage.