ADA-C01 Question 223
Single answerImplement and maintain a scaling policyA Snowflake administrator manages a multi-cluster virtual warehouse used by a BI tool during business hours. The warehouse is configured with AUTO_SUSPEND = 300, AUTO_RESUME = TRUE, MIN_CLUSTER_COUNT = 1, and MAX_CLUSTER_COUNT = 5. Users report that dashboard queries are frequently queued between 9:00 AM and 11:00 AM, but the finance team also wants to avoid unnecessary compute spend after the rush period. Which warehouse scaling policy should the administrator configure to best balance performance during peak concurrency with cost control?
- A
Set the scaling policy to STANDARD so Snowflake adds clusters more aggressively when queries start queuing and removes them less aggressively afterward.
- B
Set the scaling policy to ECONOMY so Snowflake delays adding clusters until each cluster is fully utilized, reducing queueing during peak periods.
- C
Set the scaling policy to ECONOMY because it removes clusters more quickly and also adds clusters sooner when brief queueing is detected.
- D
Keep the scaling policy unchanged, because scaling policy affects only warehouse size changes, not multi-cluster behavior.
Show answer and explanation
Correct answer: A
Explanation
The best answer is STANDARD. In Snowflake, scaling policy applies to multi-cluster warehouses and determines how aggressively additional clusters are started and stopped within the configured MIN_CLUSTER_COUNT and MAX_CLUSTER_COUNT range. STANDARD is the recommended choice when minimizing query queueing and supporting concurrency is more important, while ECONOMY is better when cost optimization is prioritized and some additional queueing is acceptable. In this scenario, users are already experiencing queueing during peak BI hours, so STANDARD is the most appropriate policy to improve responsiveness without removing overall cost controls, since MAX_CLUSTER_COUNT still limits scale-out and AUTO_SUSPEND helps reduce idle usage. This aligns with Snowflake best practices for concurrency-sensitive reporting workloads.
- A. Correct.
Correct. For a multi-cluster warehouse, the STANDARD scaling policy is designed to prioritize performance by adding clusters more proactively when demand increases and by being less aggressive about removing clusters. This helps reduce query queueing during high-concurrency periods. In this scenario, that aligns with the BI workload, where user experience during peak hours is important, while MAX_CLUSTER_COUNT still provides an upper bound on cost.
- B. Incorrect.
Incorrect. ECONOMY is intended to conserve credits by being more conservative about adding clusters and more aggressive about removing them. That can increase the likelihood or duration of queueing during peak concurrency. The option incorrectly claims it would reduce queueing during peak periods.
- C. Incorrect.
Incorrect. This reverses how ECONOMY behaves. ECONOMY does help reduce spend by removing clusters sooner, but it does not add clusters sooner than STANDARD. It is specifically less aggressive about scaling out, which can hurt performance for bursty BI concurrency.
- D. Incorrect.
Incorrect. Scaling policy applies to multi-cluster warehouse scaling behavior, specifically how Snowflake starts and stops clusters between the configured minimum and maximum cluster counts. It does not control warehouse size changes such as SMALL to MEDIUM, but it absolutely does affect multi-cluster behavior.