ML Pipeline with PAI and RDS Data Access

Scenario

A data science team configures PAI workspace roles and RAM policies so their ML training jobs can access feature data stored in an ApsaraDB RDS instance, requiring both PAI permission setup and RDS account creation with appropriate database privileges.

How the products combine

1. pai · pai-manage-permissions — Platform for AI (PAI) — Manage platform access and permissions

See pai/pai-manage-permissions.

2. rds · rds-manage-accounts — ApsaraDB RDS — Manage database accounts and permissions

See rds/rds-manage-accounts.

Typical questions

• set up PAI to read from RDS
• configure ML pipeline database access
• grant PAI workspace access to training data
• PAI访问RDS数据库配置
• ML训练任务连接RDS
• create RDS account for PAI jobs
• PAI workspace permissions for database
• end-to-end ML data pipeline setup

Q: How do I configure PAI to access RDS for ML training pipelines? A: You enable ML training jobs to access RDS by configuring PAI workspace roles and RAM policies alongside creating an RDS account with appropriate database privileges. This cross-product setup manages both platform permissions and database account creation to securely connect the services.