Skip to content

Lakehouse

David Liu edited this page Sep 5, 2025 · 1 revision

Data Lakehouse

A data management architecture

  • A unified architecture that combines data lakes and data warehouses.
  • support ACID transactions on data lake
  • enables business intelligence (BI) and machine learning (ML) on all data.
  • Introduced by Databricks

Lakehouse data is organized in a schema-on-read format

  • Define the schema as needed rather than having a predefined schema

  • ? Data harvest

Load data into a lakehouse

Microsoft: You can follow the ETL (Extract, Transform, Load) process to ingest and transform data before loading to the lakehouse.

  • Why not directly load? This is anti-pattern against data lake
  • This feature gap in Lakehouse introduce market for Fivetran, Azure Data Factory and other ingestion only solution

vendor

  • StarRocks

Ref

Clone this wiki locally