A Snowflake account has a virtual warehouse named `ANALYTICS WH' configured with multi- clustering. The MIN CLUSTER COUNT is set to 1, and the MAX CLUSTER COUNT is set to 3. During a period of heavy q

Question

Options

• ASnowflake will automatically suspend all clusters when the query load drops below a certain
• BSnowflake will automatically reduce the number of active clusters down to the
• CSnowflake will immediately suspend all idle clusters to minimize credit consumption.
• DSnowflake will maintain all three clusters indefinitely, regardless of the query load, until the
• EThe time it takes for Snowflake to reduce the number of clusters depends on the `SCALING

• A
  7% (4)
• B
  76% (45)
• C
  3% (2)
• D
  14% (8)

Explanation

B is correct because when query load decreases, Snowflake's multi-cluster warehouse automatically scales in, reducing active clusters back toward the MIN_CLUSTER_COUNT (1 in this case) - this is the core purpose of auto-scaling. E is correct because the SCALING_POLICY setting (STANDARD or ECONOMY) directly controls the aggressiveness of scale-in behavior: STANDARD removes idle clusters sooner, while ECONOMY waits longer to confirm sustained low load before reducing clusters.

Why the distractors are wrong:

• A is wrong because Snowflake never suspends all clusters - it always keeps at least MIN_CLUSTER_COUNT (1) running.
• C is wrong on two counts: scale-in is not immediate (there's a cooldown/evaluation period), and again, Snowflake won't suspend all clusters below the minimum.
• D is wrong because auto-scaling warehouses do not hold clusters indefinitely - automatic scale-in is the whole point of setting a MIN below MAX.

Memory tip: Think of MIN_CLUSTER_COUNT as your permanent floor - Snowflake will never go below it. Think of SCALING_POLICY as the "patience dial" - it controls how long Snowflake waits before deciding a cluster is truly idle and worth shutting down.

No community discussion yet for this question.