To collaborate in DataWorks, your team needs to manage project isolation, user permissions, and compute resources. Workspaces are the core organizational units designed to meet these needs. A workspace provides an isolated environment where you can organize development tasks, assign member roles, and associate compute engines, ensuring an orderly, secure, and efficient data development workflow.
Workspace overview
A workspace is the basic unit for task development and for managing member permissions in DataWorks. All development activities take place within a custom workspace. In a workspace, you can add multiple data sources to manage your data and associate various compute resources to develop tasks and run scheduled jobs. A workspace administrator adds members to the workspace and assigns them specific roles, such as workspace administrator, data analyst, deployer, developer, data governance administrator, model designer, guest, O&M, project owner, and security administrator, to enable collaboration.
Workspace modes
DataWorks provides two workspace modes: basic mode and standard mode. These modes have different development workflows and data security controls. For production environments, we recommend using standard mode. For more information, see Differences between basic and standard modes and Upgrade a workspace mode.
Core workspace operations
Actions | Description | References | |
Lifecycle management | Create a workspace | Plan and create workspaces based on different business scenarios to achieve service isolation. DataWorks provides a default workspace for new users to quickly explore the service. We strongly recommend that you do not use it for any production work. For more information about the differences between the default workspace and a custom workspace, see Default system workspace. | |
Delete a workspace | When a workspace and all its assets are no longer needed, you can delete it. Deleting a workspace permanently removes its resources to prevent potential billing charges or residual data. | ||
Disable a workspace | If you do not need a workspace temporarily but want to retain its code and configurations for future use, you can disable it. Disabling a workspace pauses task scheduling while preserving its history for reactivation. | ||
Configure workspace settings | View and modify workspace configurations, such as upgrading the workspace mode. | ||
Resource configuration | Associate compute resources | Associate compute resources, such as MaxCompute, EMR Serverless Spark, and Hologres, for task development and periodic scheduling in the workspace. When you associate a compute resource, the system automatically creates a corresponding data source. | |
Add a data source | Add data sources, such as MaxCompute, E-MapReduce, and ClickHouse, to establish connections to various data storage services. These data sources are used for task execution in DataWorks and lay the foundation for subsequent data synchronization and analysis. Important Before you add a data source, make sure you understand the differences between basic and standard modes. Your data development workflow and permission controls are configured based on the features of the selected mode. | ||
Associate a resource group | A DataWorks resource group is a unit of computing power required to run tasks. Associate a resource group with a workspace to manage the computing power for its tasks. | ||
Team and permissions | Plan and assign member roles |
Important Before you assign permissions, make sure you understand the DataWorks workspace-level permission model. | |
Production guidelines
Regardless of how you plan your workspaces, we recommend following these general guidelines:
Plan your workspaces in advance: Before creating a workspace, see Plan workspaces. Design a logical structure based on business lines, projects, or data domains to avoid management issues later.
Name your workspaces clearly: Use names that reflect the business purpose, such as
finance_tax_report.If you reference nodes across workspaces, we recommend including an abbreviation of the workspace name in the node name for easy identification.
Choose the right workspace mode: Standard mode provides physical isolation between the development environment and the production environment. We recommend using standard mode for production.
Manage workspace permissions: Workspaces should be created by an Alibaba Cloud account. Then, assign the Workspace Administrator role to RAM users (sub-accounts). As a best practice, a single user should not be assigned both the development and O&M roles.
Avoid using the default workspace for production: The default workspace is not recommended for production work.
Next steps
After you create a workspace and assign member roles, you can start Data Integration, data development, or data analysis based on your business needs.