SnowPro Associate: Platform Question 255
Single answer○ Scaling up or downA BI team uses a Snowflake virtual warehouse named BI_WH, sized MEDIUM, to run dashboard queries during business hours. Users report that individual dashboard queries are taking too long to finish, even though there are only a few concurrent users at a time. The administrator wants to improve the performance of each query as quickly as possible without changing the SQL. Which action should the administrator take?
- A
Resize BI_WH from MEDIUM to LARGE
- B
Enable multi-cluster on BI_WH with a maximum of 3 clusters
- C
Create a resource monitor to suspend BI_WH when credits exceed the daily quota
- D
Convert BI_WH to a Snowpark-optimized warehouse
Show answer and explanation
Correct answer: A
Explanation
When users need faster performance for individual queries and concurrency is low, the best action is typically to scale up the warehouse size. In Snowflake, resizing a virtual warehouse increases the compute resources available to execute queries. By contrast, scaling out with multi-cluster warehouses is intended to reduce queueing and support more concurrent workloads, not primarily to accelerate a single query. Resource monitors are for cost control, and Snowpark-optimized warehouses are specialized for memory-intensive processing rather than standard BI reporting. This aligns with Snowflake best practices for warehouse sizing: scale up to improve single-query performance, and scale out to address concurrency.
- A. Correct.
Correct. Increasing the warehouse size from MEDIUM to LARGE scales up the compute resources available to each query. This is the appropriate action when the problem is query execution speed and the workload has low concurrency. Scaling up gives individual queries more CPU, memory, and processing capacity, which can reduce runtime for many workloads.
- B. Incorrect.
Incorrect. Multi-cluster warehouses are primarily used to handle higher concurrency by adding additional clusters when many queries are queued at the same time. In this scenario, only a few users are running queries concurrently, so the issue is not concurrency. Enabling multi-cluster would increase the ability to process more simultaneous queries, but it would not be the best first step to speed up each individual query.
- C. Incorrect.
Incorrect. A resource monitor helps control or monitor credit consumption, for example by sending notifications or suspending a warehouse when usage thresholds are reached. It does not improve query performance or make individual queries run faster. This option addresses cost governance, not performance tuning.
- D. Incorrect.
Incorrect. Snowpark-optimized warehouses are designed for specific memory-intensive workloads, such as certain Snowpark and machine learning use cases. Standard BI dashboard queries do not typically require a Snowpark-optimized warehouse, and changing to this warehouse type is not the normal solution for improving general SQL dashboard performance.