You are viewing a plain text version of this content. The canonical link for it is here.

Posted to commits@dolphinscheduler.apache.org by "eye-gu (via GitHub)" <gi...@apache.org> on 2026/06/24 05:16:35 UTC

[I] [Improvement][datasource] Enable users to customize datasources [dolphinscheduler]

eye-gu opened a new issue, #18377:
URL: https://github.com/apache/dolphinscheduler/issues/18377

### Search before asking

- [x] I had searched in the [issues](https://github.com/apache/dolphinscheduler/issues?q=is%3Aissue) and found no similar feature requirement.

### Description

## Background

DolphinScheduler's datasource layer is already plugin-based via SPI. However, datasource plugins still cannot be extended by users with their own custom types. The root cause is that the DbType enum is hardwired into multiple layers of the type system, making it physically impossible for an external plugin to register a type that doesn't exist in the built-in enum.

1. The DbType enum is hardcoded throughout the SPI and storage layers

The DataSourceProcessor.getDbType() method returns DbType (a compile-time enum). An external plugin jar cannot return a DbType constant that doesn't exist in dolphinscheduler-spi, so it can only override existing types, never add new ones.

2. UI datasource type lists are hardcoded in the frontend

The "create datasource" dialog and the task-node datasource selector both use hardcoded type arrays. Even if a custom plugin is deployed, its type never appears in the UI. The lists should be fetched from a backend API (e.g. GET /datasources/types) that returns all registered type names. For custom/unknown types, the generic JDBC form (host/port/user/password/database) is sufficient for the vast majority of cases.

3. t_ds_datasource.type column needs to change from numeric code to string name

Currently tinyint storing the enum's integer code. It should become varchar storing the type name string (e.g. "MYSQL"). A migration is needed to translate existing integer codes to name strings.The task definition / task instance tables already store the datasource type as a string name inside the task_params JSON (e.g. "type":"MYSQL"), so no change is needed there.

## Proposed Solution

Relax DbType to a built-in constant collection (keep the enum, remove the integer code). Change all type-lookup and storage paths to use String type names — aligned with the existing task plugin design (TaskDefinition.taskType is already String).
Add a GET /datasources/types endpoint returning all registered type names. Frontend fetches this for both the datasource dialog and task-node dropdowns, falling back to a generic JDBC form for non-builtin types.
Change t_ds_datasource.type to varchar(64)

### Are you willing to submit a PR?

- [x] Yes I am willing to submit a PR!

### Code of Conduct

- [x] I agree to follow this project's [Code of Conduct](https://www.apache.org/foundation/policies/conduct)

--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@dolphinscheduler.apache.org.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org