Apache Hudi

Open-source data lake framework providing streaming data lake capabilities with ACID transactions, upserts, and incremental processing.

Visit Website → GitHub

Key Features

✓ Upserts: Insert or update based on key
✓ Deletes: Soft and hard deletes
✓ Merge on read: Lazy reconciliation
✓ Copy on write: Immediate consistency
✓ Near real-time: Minutes latency
✓ Kafka integration: Direct streaming ingest
✓ DeltaStreamer: Built-in ingestion utility
✓ Flink support: Stream processing integration
✓ Change streams: Track data changes
✓ Incremental queries: Process only deltas
✓ Time travel: Query historical versions
✓ Rollback: Revert to previous state
✓ Atomicity: All-or-nothing commits
✓ Consistency: Schema enforcement
✓ Isolation: Concurrent reader/writer
✓ Durability: Committed data persists