After nearly a year of iterative optimization, LakeSoul 3.0.0 is officially released. This release brings the following key updates:

This presentation explains LakeSoul's Rust-native IO architecture and its integration with compute engines.

This presentation introduces the LakeSoul project, its architecture, and its core use cases.

Recently, LakeSoul, the lakehouse framework released version 2.3.0. This new release is the first release of LakeSoul after it entered the incubation of the Linux Foundation AI & Data as a sandbox project. This new version adds Flink SQL/Table API, which supports stream and batch read and write. The Flink DataStream API for multi-table real-time CDC stream ingestion has been refactored to better support data ingestionfrom multiple data sources to the lakehouse. A new global automatic small file compaction service has been added.

Recently, after months of research and development, LakeSoul released version 2.2.0(Github Release Notes). The most important upgrade in this version is that the new Native IO is enabled by default in both Spark and Flink, and LakeSoul's performance has once again significantly been improved and expanded its performance leadership advantage in the field of Cloud Native Data Lakehouse.This article provides you with a detailed explanation of the updates to LakeSoul version 2.2.0 and the technical details of Native IO.