ADA-C01 Question 404
Single answerLatency for historical viewsA Snowflake administrator is troubleshooting a monitoring dashboard that relies on ACCOUNT_USAGE historical views to report warehouse credit consumption and query activity. Operations complains that the dashboard often appears incomplete for the most recent time period, especially after workload spikes. The administrator needs to set correct expectations and reduce false incident escalations. Which action is the MOST appropriate?
- A
Continue using ACCOUNT_USAGE views for near-real-time alerting, because these views are updated immediately after each query completes
- B
Switch the dashboard to ORGANIZATION_USAGE views, because they have lower latency than ACCOUNT_USAGE for recent operational monitoring
- C
Redesign the dashboard to account for ACCOUNT_USAGE view latency and use lower-latency Information Schema table functions where near-real-time visibility is required
- D
Enable Search Optimization Service on the SNOWFLAKE database so historical views refresh faster for recent rows
Show answer and explanation
Correct answer: C
Explanation
The key concept is that Snowflake historical views, especially those in SNOWFLAKE.ACCOUNT_USAGE, are not guaranteed to be real-time. They are widely used for administration, governance, and cost analysis, but administrators must understand and communicate their documented latency characteristics. When teams need fresher operational visibility, Snowflake documentation generally points to Information Schema sources, including relevant table functions for active and recent activity, as the better fit. In practice, the right solution is not to force ACCOUNT_USAGE into a real-time monitoring role, but to design dashboards and alerting around the intended freshness of each metadata source. This is a common SnowPro Advanced Administrator scenario: choosing the correct metadata interface based on latency and operational requirements rather than relying on a single historical source for every use case.
- A. Incorrect.
Incorrect. ACCOUNT_USAGE views are historical views in the shared SNOWFLAKE database and are not intended for immediate, real-time monitoring. Many of these views have documented latency before new records appear. Using them for near-real-time alerting commonly causes missing or incomplete recent results, which is exactly the issue in the scenario.
- B. Incorrect.
Incorrect. ORGANIZATION_USAGE is also intended for historical reporting and governance across accounts, not lower-latency operational monitoring. It does not solve the recent-data delay problem and may introduce additional scope or access considerations rather than improving freshness.
- C. Correct.
Correct. This is the best administrative response. ACCOUNT_USAGE views are excellent for historical analysis, trending, auditing, and chargeback, but they can have latency before data becomes visible. For operational dashboards or alerting that require fresher data, Snowflake best practice is to use the appropriate Information Schema table functions or views where available, since they generally provide lower-latency access to current activity. The dashboard should also clearly communicate that historical views may lag.
- D. Incorrect.
Incorrect. Search Optimization Service improves selective query performance on user tables in supported scenarios; it does not change how quickly Snowflake populates metadata or historical views in the SNOWFLAKE shared database. This option reflects a common misconception between query performance optimization and metadata refresh latency.