Release Notes — June 16, 2026 (Lakehouse Studio v2.2.1)

This release (Release 2026.06.16, v2.2.1) introduces new features, enhancements, and fixes. These updates will roll out in stages to the following regions. Updates complete within one to two weeks of the release date, depending on your region.

Domestic regions

Alibaba Cloud (Shanghai)
Tencent Cloud (Shanghai / Beijing / Guangzhou)
AWS (Beijing)
Alibaba Cloud (Hangzhou)

International regions

Alibaba Cloud (Singapore)
AWS (Singapore)

Breaking Changes

Compute cluster: the shared cluster has been taken offline; historical data has been corrected.

AI Capabilities

DataAgent: The Data Engineering Agent is now generally available. This AI agent covers the full cycle of data development, operations, and governance. Unlike the traditional "humans write code, tools execute" model, DataAgent lets users express intent in natural language while the Agent handles modeling, development, scheduling, and diagnosis — enabling a true "humans direct Agents" collaboration pattern. Whether building a data warehouse from scratch, diagnosing task failures, or running ad-hoc analysis, DataAgent can take on the work and significantly lower the entry bar and reduce repetitive effort in data engineering. Data Engineering Agent

New Features

Platform-wide: Workspace-level global parameters are now supported. Previously, environment configurations, file paths, dates, and similar variables were scattered across individual tasks, requiring repeated maintenance and prone to errors. You can now define parameters centrally in a workspace, reuse them across tasks, and have a single change take effect everywhere — making environment switching between development, test, and production cleaner and more controllable. Workspace Global Parameters
Notebook: This release delivers an overall performance optimization for Notebook, significantly improving code execution response speed and run stability in large-data scenarios, making interactive development smoother. Additionally, Notebook now supports the Spark cluster type, enabling interactive development and debugging on the Spark engine for large-scale data processing and complex transformations that SQL cannot easily cover.
Data Integration: Offline integration tasks for StarRocks now support checkpoint resume, and the bitmap and array data types are supported.
Task Development: Task orchestration now includes a Merge node, which consolidates multiple execution paths split by a branch node back into a unified downstream node. Used together with branch nodes, this more elegantly expresses "split by condition, then converge" orchestration logic, making task flow diagrams clearer and dependency relationships easier to maintain. Merge Task
Data Quality: Improved DQC error log details to surface failure reasons more quickly and clearly.

Known Limitations

None