Multiple analysts run frequent ad-hoc SQL against curated Delta tables throughout the day. Requirements: fast startup, high concurrency for many simultaneous users, automatic scaling to control cost, and no need to manage Spark cluster internals. Which compute option is the best fit?
Answer & explanation
Correct answer: B
WHY B: A serverless SQL warehouse is purpose-built for interactive, high-concurrency SQL: near-instant startup, autoscaling across concurrent queries, and no cluster management, which matches every requirement.
WHY NOT A: A single-node cluster cannot serve many concurrent analysts and carries a slow, manual lifecycle.
WHY NOT D: Spinning up a job cluster per query adds minutes of startup latency and cost and is not how interactive SQL is served.
WHY NOT C: A fixed, always-on large cluster wastes money (no scaling to control cost) and still requires manual configuration and management.
WHY NOT E: Structured Streaming is for continuous pipelines, not ad-hoc interactive SQL.