ADA-C01 Question 215
Single answerConfigure warehouse properties (auto-suspend, auto-resume, max clusters)A Snowflake administrator manages a reporting warehouse used by BI dashboards during business hours and by a small number of ad hoc analysts throughout the day. Users report that dashboard queries are occasionally queuing during peak morning usage, but leadership also wants to minimize compute spend when the warehouse is idle. The administrator wants the warehouse to start automatically when users submit queries, suspend shortly after activity stops, and scale out only when concurrency increases. Which warehouse configuration best meets these requirements?
- A
Configure the warehouse with AUTO_SUSPEND = 60, AUTO_RESUME = TRUE, MIN_CLUSTER_COUNT = 1, and MAX_CLUSTER_COUNT = 3
- B
Configure the warehouse with AUTO_SUSPEND = 0, AUTO_RESUME = FALSE, MIN_CLUSTER_COUNT = 3, and MAX_CLUSTER_COUNT = 3
- C
Configure the warehouse with AUTO_SUSPEND = 600, AUTO_RESUME = TRUE, and MAX_CLUSTER_COUNT = 1
- D
Configure the warehouse with AUTO_SUSPEND = 60, AUTO_RESUME = TRUE, MIN_CLUSTER_COUNT = 3, and MAX_CLUSTER_COUNT = 1
Show answer and explanation
Correct answer: A
Explanation
For this scenario, the best practice is to combine automatic start/stop behavior with multi-cluster scaling for concurrency. AUTO_RESUME should be enabled so users and BI tools do not need to manually start the warehouse. A short AUTO_SUSPEND value helps control cost when the warehouse is idle. To address queuing caused by concurrent dashboard queries, the warehouse should be configured as a multi-cluster warehouse with MIN_CLUSTER_COUNT set low enough to avoid unnecessary baseline cost and MAX_CLUSTER_COUNT set higher to allow scale-out during peak concurrency. Snowflake documentation describes AUTO_SUSPEND and AUTO_RESUME as core cost-management settings and explains that multi-cluster warehouses help with concurrency by adding clusters up to MAX_CLUSTER_COUNT as needed.
- A. Correct.
Correct. AUTO_SUSPEND = 60 helps reduce cost by suspending the warehouse after a short idle period. AUTO_RESUME = TRUE ensures queries can start the warehouse automatically without manual intervention. Setting MIN_CLUSTER_COUNT = 1 keeps a single cluster running under normal load, while MAX_CLUSTER_COUNT = 3 allows a multi-cluster warehouse to add clusters during periods of high concurrency, which helps reduce query queuing for BI workloads.
- B. Incorrect.
Incorrect. AUTO_SUSPEND = 0 effectively disables automatic suspension, which conflicts with the requirement to minimize spend when idle. AUTO_RESUME = FALSE also conflicts with the requirement for the warehouse to start automatically when users submit queries. Keeping both minimum and maximum clusters at 3 would force three clusters to run continuously, increasing cost even during low usage.
- C. Incorrect.
Incorrect. AUTO_RESUME = TRUE is appropriate, but AUTO_SUSPEND = 600 keeps the warehouse running for 10 minutes after activity stops, which is less cost-efficient than a shorter suspend period for intermittent usage. More importantly, MAX_CLUSTER_COUNT = 1 means the warehouse cannot scale out to additional clusters, so it will not address queuing caused by high concurrency.
- D. Incorrect.
Incorrect. This option is internally inconsistent for the stated goal. In Snowflake, the maximum cluster count must be greater than or equal to the minimum cluster count. Setting MIN_CLUSTER_COUNT = 3 and MAX_CLUSTER_COUNT = 1 is not a valid way to configure a warehouse, and even conceptually it would not support cost-efficient scaling behavior.