DAA-C01 Question 61
Single answerUse logging and monitoring solutionsA data analyst supports several dashboards that run on a shared virtual warehouse. Business users report that one dashboard becomes slow every morning between 8:00 AM and 9:00 AM, but only on weekdays. The analyst needs to identify whether the slowdown is caused by warehouse queuing, poor query performance, or changes in workload patterns, and wants a repeatable way to monitor this issue over time inside Snowflake. Which solution would BEST meet these requirements?
- A
Query the ACCOUNT_USAGE.QUERY_HISTORY view and compare average execution time by hour of day for the dashboard queries only.
- B
Use Snowsight's Query History and Warehouse Activity/Monitoring views to inspect queued time, load, and concurrency for the affected period, then create repeatable monitoring queries against ACCOUNT_USAGE views such as QUERY_HISTORY and WAREHOUSE_LOAD_HISTORY.
- C
Review the database's Time Travel history to determine whether data retention settings are causing the morning slowdown, and schedule alerts from there.
- D
Increase the warehouse size immediately and rely on the dashboard tool's refresh duration logs instead of Snowflake monitoring views.
- E
Use COPY_HISTORY to determine whether dashboard queries are waiting on file ingestion and create a recurring report from that view.
Show answer and explanation
Correct answer: B
Explanation
The best answer is to combine Snowflake's built-in monitoring interfaces in Snowsight with repeatable analysis using ACCOUNT_USAGE views. In practice, a data analyst would first inspect the affected time window in Snowsight to quickly validate whether queries are queuing and whether warehouse load or concurrency spikes during weekday mornings. For repeatable monitoring, Snowflake's ACCOUNT_USAGE schema provides views such as QUERY_HISTORY for query-level performance analysis and WAREHOUSE_LOAD_HISTORY for warehouse-level load and queuing trends. This aligns with Snowflake best practices: use query history to analyze execution characteristics, and use warehouse monitoring/load history to determine whether contention or concurrency scaling issues are affecting performance. Time Travel and COPY_HISTORY address different operational concerns and do not solve this monitoring use case. A reactive warehouse resize without root-cause analysis is also weaker than using Snowflake logging and monitoring capabilities to diagnose and track the issue over time.
- A. Incorrect.
This is incomplete. QUERY_HISTORY is useful for identifying slow queries and comparing execution characteristics over time, but by itself it does not directly provide the clearest warehouse-level view of queuing, concurrency, and overall warehouse load patterns. The scenario specifically asks to distinguish among queuing, query inefficiency, and workload changes, and to create repeatable monitoring. QUERY_HISTORY alone may help with elapsed time and some query-level metrics, but combining it with warehouse monitoring data is the stronger and more complete approach.
- B. Correct.
This is correct. Snowsight provides built-in monitoring views for query activity and warehouse performance, which are appropriate for an initial investigation during the 8:00-9:00 AM window. For ongoing monitoring, ACCOUNT_USAGE views such as QUERY_HISTORY and WAREHOUSE_LOAD_HISTORY are designed for historical analysis of query performance, queuing, and warehouse load trends. This combination helps determine whether the slowdown is caused by warehouse contention, increased concurrency, or slower individual queries, and it supports a repeatable internal monitoring solution in Snowflake.
- C. Incorrect.
This is incorrect. Time Travel is for accessing historical table data, not for performance monitoring of dashboards or warehouse queuing analysis. Retention settings do not serve as a primary mechanism to diagnose recurring one-hour dashboard slowdowns. Someone might choose this option because it involves 'history,' but it is the wrong type of history for operational monitoring.
- D. Incorrect.
This is incorrect. Increasing warehouse size may reduce some performance issues, but doing so before diagnosing the cause is not a best practice. The scenario explicitly requires identifying whether the issue is queuing, query performance, or workload pattern changes. External dashboard refresh logs may show symptoms, but Snowflake monitoring data is the authoritative source for understanding warehouse load, query queuing, and execution behavior.
- E. Incorrect.
This is incorrect. COPY_HISTORY tracks data loading activity, especially COPY INTO operations, not dashboard query execution or warehouse queuing for BI workloads. A candidate might confuse ingestion monitoring with query monitoring, but the problem is about dashboard slowness during a recurring business hour window, which is better analyzed with query and warehouse usage views.