COF-C03 Question 203
Single answerCost and warehouse monitoringA Snowflake administrator notices that warehouse credit consumption increased sharply over the last two weeks, even though the number of scheduled ETL jobs has not changed. The team wants to identify which warehouses are driving the increase and determine whether query execution patterns or warehouse sizing are contributing to the cost spike. Which Snowflake feature should the administrator use first to get the most direct account-level view of warehouse consumption trends and support this investigation?
- A
Query the WAREHOUSE_METERING_HISTORY view in the ACCOUNT_USAGE schema
- B
Query the LOGIN_HISTORY view in the ACCOUNT_USAGE schema
- C
Review only the QUERY_HISTORY view to calculate warehouse credits consumed per query
- D
Use the STORAGE_USAGE view in the ACCOUNT_USAGE schema to identify which warehouse is consuming the most credits
Show answer and explanation
Correct answer: A
Explanation
For cost and warehouse monitoring, the best first step is to analyze warehouse metering data using SNOWFLAKE.ACCOUNT_USAGE.WAREHOUSE_METERING_HISTORY. This view is designed to show compute credit usage by warehouse over time, making it the most appropriate source for identifying cost increases tied to specific warehouses. Once high-cost warehouses are identified, administrators can further investigate with QUERY_HISTORY, WAREHOUSE_LOAD_HISTORY, and warehouse configuration details such as size, auto-suspend, and auto-resume settings. This aligns with Snowflake best practices for monitoring compute costs: start with account-level metering, then drill into workload and tuning details. Relevant Snowflake documentation includes ACCOUNT_USAGE views such as WAREHOUSE_METERING_HISTORY, QUERY_HISTORY, and WAREHOUSE_LOAD_HISTORY, which together support cost analysis and warehouse performance investigation.
- A. Correct.
Correct. WAREHOUSE_METERING_HISTORY in SNOWFLAKE.ACCOUNT_USAGE is the primary account-level source for analyzing warehouse credit consumption over time. It allows administrators to see metering by warehouse and time period, which is the most direct starting point for identifying which warehouses are responsible for increased spend. After identifying the warehouses with higher consumption, the administrator can correlate that information with query patterns, warehouse size, auto-suspend behavior, and concurrency.
- B. Incorrect.
Incorrect. LOGIN_HISTORY tracks user authentication events, such as successful and failed logins. While it may help investigate user access or suspicious account activity, it does not provide warehouse credit consumption details and is not the right source for warehouse cost analysis.
- C. Incorrect.
Incorrect. QUERY_HISTORY is useful for investigating workload patterns, long-running queries, and warehouse usage context, but it is not the best first source for direct warehouse credit consumption trends. Credit consumption is metered at the warehouse level, and WAREHOUSE_METERING_HISTORY provides that information more directly. A common misconception is that per-query analysis should come first, but administrators typically start with warehouse metering and then drill into query history if needed.
- D. Incorrect.
Incorrect. STORAGE_USAGE is used for analyzing storage-related costs, not virtual warehouse compute credits. Warehouses consume compute credits, so this view would not identify which warehouse is driving the compute cost increase. This distractor reflects confusion between storage charges and compute charges in Snowflake.