Reviewed by Aditya Kumar · Last reviewed 2026-03-24
**Why DBU understanding matters**: Databricks bills by DBU (Databricks Unit)—normalized compute metric. Cost = DBU rate × DBU consumption × runtime. **DBU by workload**: All-purpose clusters vs job clusters have different rates; Photon has different DBU multipliers. Serverless...
This hard-level Spark/Big Data question appears frequently in data engineering interviews at companies like JP Morgan. While less common, it tests deeper understanding that distinguishes strong candidates. Mastering the underlying concepts (etl, optimization, sql) will help you answer variations of this question confidently.
This is a senior-level question that tests architectural thinking. Lead with the high-level design, then drill into specifics. Discuss trade-offs explicitly - there is rarely one correct answer. Show awareness of scale, fault tolerance, and operational complexity.
Why DBU understanding matters: Databricks bills by DBU (Databricks Unit)—normalized compute metric. Cost = DBU rate × DBU consumption × runtime. DBU by workload: All-purpose clusters vs job clusters have different rates; Photon has different DBU multipliers. Serverless has different pricing. Optimization levers: Right-size clusters (avoid over-provisioning); use spot instances for batch; autoscaling; Photon for SQL/ETL (often 2–3x faster, similar or lower DBU); scale-to-zero for job clusters. Scalability trade-offs: Larger clusters = higher DBU but faster—optimize total cost = DBU × time, not just cluster size. Cost implications: Job clusters cheaper than all-purpose; spot can cut 50–70%; idle clusters cost money. Best practice: Monitor usage by workload; use job clusters for batch; enable Photon for compatible workloads.
Red Flag: Leaving all-purpose clusters running 24/7. Pro-Move: 'We use job clusters with spot; Photon = 40% DBU savings on SQL workloads.'
Some links below are affiliate links. If you buy through them we may earn a small commission at no extra cost to you — it helps keep DataEngPrep free.
According to DataEngPrep.tech, this is one of the most frequently asked Spark/Big Data interview questions, reported at 1 company. DataEngPrep.tech maintains an editor-reviewed database of 1,863 data engineering interview questions across 7 categories.