ADA-C01 Question 419
Single answerMonitor and assess automatic clustering usageA Snowflake administrator notices a sharp increase in monthly credits for a production database after several large batch MERGE operations were introduced. The largest table in the database has a clustering key defined, and the team suspects Automatic Clustering is contributing to the increase. The administrator needs to determine whether Automatic Clustering is being used heavily on this table and assess its impact over time. Which approach should the administrator take?
- A
Query the ACCOUNT_USAGE.AUTOMATIC_CLUSTERING_HISTORY view to review credits consumed and rows/bytes reclustered for the table over time.
- B
Query the ACCOUNT_USAGE.QUERY_HISTORY view and filter for reclustering SQL statements generated by the Automatic Clustering service.
- C
Use SHOW TABLES to compare the current clustering key with the table's average depth and infer total Automatic Clustering credits consumed.
- D
Query WAREHOUSE_METERING_HISTORY for the virtual warehouse used by ETL jobs, because Automatic Clustering always runs on customer-managed warehouses.
Show answer and explanation
Correct answer: A
Explanation
To monitor and assess Automatic Clustering usage, administrators should use Snowflake's usage views designed specifically for Snowflake-managed services. The ACCOUNT_USAGE.AUTOMATIC_CLUSTERING_HISTORY view is the key source for evaluating when Automatic Clustering ran, how much it consumed, and which objects were affected. This is especially important after heavy DML operations such as MERGE, UPDATE, or DELETE, which can increase reclustering work on clustered tables. QUERY_HISTORY can help correlate workload changes that caused reclustering pressure, but it is not the primary source for Automatic Clustering service consumption. WAREHOUSE_METERING_HISTORY is also not appropriate because Automatic Clustering is a serverless Snowflake-managed service rather than a workload executed on a customer warehouse. Best practice is to combine object-level Automatic Clustering history with an understanding of DML patterns and clustering-key design to determine whether the clustering strategy remains cost-effective. Relevant Snowflake documentation includes the Automatic Clustering monitoring guidance and the ACCOUNT_USAGE views for serverless feature usage.
- A. Correct.
Correct. ACCOUNT_USAGE.AUTOMATIC_CLUSTERING_HISTORY is the appropriate source to monitor and assess Automatic Clustering usage. It provides historical usage details such as credits consumed and reclustering activity, allowing the administrator to analyze whether a specific table is incurring significant Automatic Clustering cost after large DML changes such as MERGE operations.
- B. Incorrect.
Incorrect. Automatic Clustering is a Snowflake-managed service, not a user-submitted SQL workload that appears as ordinary reclustering statements in QUERY_HISTORY. While QUERY_HISTORY is useful for analyzing user queries and DML patterns that may trigger reclustering needs, it is not the authoritative source for measuring Automatic Clustering consumption.
- C. Incorrect.
Incorrect. SHOW TABLES and clustering metadata can help identify whether a table has a clustering key and may provide context about clustering state, but they do not provide historical credit consumption for Automatic Clustering. Inferring total service cost from table metadata alone is not reliable.
- D. Incorrect.
Incorrect. WAREHOUSE_METERING_HISTORY tracks credit usage for virtual warehouses. Automatic Clustering does not run on customer-managed warehouses, so its usage is not measured there. This option reflects a common misconception that all compute in Snowflake is tied to a warehouse.