Reviewed by Aditya Kumar · Last reviewed 2026-03-24
Strategies: (1) Spot for task nodes; On-Demand for master/core if needed. (2) Right-size instance types. (3) Auto-scaling. (4) Transient clusters—run job, terminate. (5) Spot blocks for 1–6h jobs. (6) Savings Plans. Trade-off: Spot interruptions; design for fault tolerance....
Strategies: (1) Spot for task nodes; On-Demand for master/core if needed. (2) Right-size instance types. (3) Auto-scaling. (4) Transient clusters—run job, terminate. (5) Spot blocks for 1–6h jobs. (6) Savings Plans. Trade-off: Spot interruptions; design for fault tolerance. Best practice: Spot for fault-tolerant; monitor interruption; bootstrap for fast startup.
Red Flag: On-Demand for all nodes when Spot would work. Pro-Move: 'Task nodes 100% Spot; diversified types—70% cost cut; we checkpoint for interruptions.'
According to DataEngPrep.tech, this is one of the most frequently asked Cloud/Tools interview questions, reported at 1 company. DataEngPrep.tech maintains an editor-reviewed database of 1,863 data engineering interview questions across 7 categories.