Change control

更新时间: 2026-01-18 19:09:00

The change control feature checks changes before they are executed. These checks occur during stages such as submitting, publishing, and operations and maintenance (O&M). You configure control rules and their effective scopes. The system then validates changes against these rules. The results determine if a change can proceed. This process prevents high-risk operations, such as unexpected data changes or accidental deletions, from affecting downstream services. Change control consists of two parts: change rules and change policies:

  • A change rule is a validation rule based on the metadata from an object's basic properties. It combines built-in system rules with logic from custom code. Objects that fail a rule check are handled according to the change policy.

  • A change policy defines the effective scope and control method for your change rules. You can configure multiple rules in a single policy and assign a different control method to each rule.

Dataphin supports a publish control feature that checks nodes before they are published. The checks use the control rules and policies that you set. The results determine whether a node meets the publishing conditions. For nodes that fail a check, the system either blocks the publishing flow (strong control) or displays a warning without blocking the publish (weak control), depending on the rule's configuration. This enhances platform stability.

Common scenarios for publish control include the following:

  • Time-based control: Prevents publishing new nodes or updating existing ones at specific times. This avoids issues caused by changes in data calculation logic. Examples include year-end closing for a fiscal year, financial audits, control during Double 11 sales promotions, and ensuring stability during cluster switches.

  • Development standard control: Makes decisions based on node properties. For example, nodes in an operation data store (ODS) project cannot be published if their names do not start with the prefix ods_.

  • Account control: Restricts specific users from submitting or publishing nodes. For example, nodes submitted by former employees cannot be published.

After you configure the rules, you can use switches to enable or disable rules and policies. This lets you adjust settings for different scenarios, which reduces manual work and improves the platform's standards and stability.

上一篇: Sensitive data protection 下一篇: Data backup and deletion
阿里云首页 智能数据建设与治理 Dataphin 相关技术圈