Reviewed by Aditya Kumar · Last reviewed 2026-03-24
**Why Delta matters**: Parquet + transaction log = mutable lakehouse with reliability. **ACID**: Transaction log (_delta_log/) records commits; concurrent reads/writes serialized. **Time travel**: Query by version or timestamp; `VERSION AS OF n` or `TIMESTAMP AS OF '...'`....
Why Delta matters: Parquet + transaction log = mutable lakehouse with reliability. ACID: Transaction log (_delta_log/) records commits; concurrent reads/writes serialized. Time travel: Query by version or timestamp; VERSION AS OF n or TIMESTAMP AS OF '...'. Streaming: Read/write as stream; merge support; exactly-once with checkpoint. Scalability trade-offs: Log grows; VACUUM removes old files. Time travel retention = storage. Cost implications: Delta = storage + log; VACUUM reduces cost but limits time travel. Best practice: OPTIMIZE + VACUUM; set retention; use time travel for audits.
Red Flag: Not knowing VACUUM deletes old files. Pro-Move: 'We retain 7 days; VACUUM weekly; time travel for incident investigation.'
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.