How DataWorks Scaled Its IDE with Micro‑Frontend and Plugin Architecture

Overview

DataWorks is an all‑in‑one big‑data intelligent cloud development platform that provides data integration, development, governance, security, services, application development, and machine‑learning capabilities.

Key Challenges

Complex product functionality and architecture – the IDE‑style rich‑interactive single‑page application contains hundreds of nodes and modules, leading to build times over ten minutes and difficult maintenance.

Complex runtime environments – the platform must run on private clouds, public clouds, and hybrid clouds without CDN support, requiring code that can be reused across all environments.

Highly customized commercial demands – per‑user feature toggles generate massive permission‑checking code, increasing development and testing effort.

Limited gray‑release mechanisms – current manual feature‑switches are costly and error‑prone, hindering iterative testing.

Resource constraints – a small front‑end team (10+ engineers) struggles to keep up with rapid feature iteration.

Architecture Evolution

Monolithic Era

The early DataStudio IDE was built as a single large front‑end application, accumulating code and dependencies, which caused heavy deployment, long compile times, and difficult regression testing.

Componentization

Adopting a service‑oriented approach, the team extracted reusable UI components (terminal, file tree, tab panel, LSP editor) and modularized business logic, but dependency complexity increased.

Micro‑Frontend & Plugin

To address large‑scale aggregation, DataWorks introduced a micro‑frontend framework based on qiankun and single‑spa, extending it with multi‑instance rendering, slot mechanisms, style and version isolation, and nested plugins.

Plugins encapsulate UI and business logic, allowing independent development, deployment, and runtime. A slot plugin can host child plugins, enabling fine‑grained composition of toolbar buttons and other UI elements.

Unified Studio Base

The unified studio base abstracts common IDE capabilities (layout, communication, i18n, theming, event/message centers) and provides a plug‑in API for higher‑level business features. This base is shared across all vertical studios (StreamStudio, HoloStudio, GraphStudio, etc.), reducing duplication.

Plugin Management Console

A visual console lets teams configure plugin dependencies and orchestrate multi‑plugin pages without code changes, supporting both front‑end micro‑apps and back‑end micro‑services.

Future Outlook

DataWorks aims to make the entire platform micro‑frontend and plugin‑driven, enabling industry‑specific extensions, open ecosystem participation, and seamless deployment across private, public, and hybrid clouds.