DEA-C02 Question #31: Real Exam Question with Answer & Explanation

Question

The following chart represents the performance of a virtual warehouse over time: A Data Engineer notices that the warehouse is queueing queries. The warehouse is size X-Small, the minimum and maximum cluster counts are set to 1, the scaling policy is set to standard, and auto-suspend is set to 10 minutes. How can the performance be improved?

Options

• AChange the cluster settings.
• BIncrease the size of the warehouse.
• CChange the scaling policy to economy.
• DChange auto-suspend to a longer time frame.

Explanation

Option A is correct because the root cause of query queuing is a concurrency bottleneck, and the warehouse has multi-cluster warehousing effectively disabled - both min and max cluster counts are set to 1. Changing the cluster settings (specifically raising the maximum cluster count above 1) allows Snowflake to spin up additional compute clusters to handle simultaneous queries in parallel, directly resolving the queue.

Option B (increase warehouse size) is a common trap: scaling up from X-Small to a larger size improves individual query speed by adding more nodes per cluster, but it does nothing to increase concurrency - queued queries still wait in line. Option C (economy scaling policy) would make things worse, not better, since economy mode is more conservative about spinning up new clusters and waits longer before doing so. Option D (longer auto-suspend) controls how long an idle warehouse keeps running to avoid cold-start latency - it has no bearing on query queuing.

Memory tip: Queue = concurrency problem = needs more clusters, not bigger clusters. When you see queries waiting, think "horizontal" (add clusters) not "vertical" (add size).

No community discussion yet for this question.