Skip to content

[Feature] Support Apache Cloudberry as a data source #183

Description

@Arlowen

Issue type

Feature request

Summary

Please add Apache Cloudberry as a first-class data source in CloudDM.

Apache Cloudberry is an open-source MPP database derived from Greenplum and built on a modern PostgreSQL kernel. CloudDM already supports PostgreSQL and Greenplum, so their existing drivers and metadata implementations may provide a useful foundation, but Cloudberry should be identified and exposed as a separate data source so that its version differences and MPP-specific features can be handled correctly.

Actual behavior

Apache Cloudberry is not available in the data source type list. Users cannot configure it explicitly or rely on CloudDM to correctly handle Cloudberry-specific metadata, SQL syntax, and distributed table definitions.

Expected behavior

Users can add an Apache Cloudberry cluster through its coordinator endpoint and use the same core database-management workflows available for other supported relational databases.

Acceptance criteria

  • Apache Cloudberry is available as a distinct data source type with localized name, icon, and connection configuration.
  • CloudDM can connect to the Cloudberry coordinator using the PostgreSQL JDBC protocol and test the connection.
  • CloudDM detects and displays the Cloudberry product name and version correctly instead of treating it indistinguishably as PostgreSQL or Greenplum.
  • Users can browse common metadata, including databases, schemas, tables, columns, indexes, views, sequences, functions/procedures, users, and roles.
  • The SQL console supports query execution, transaction handling, SQL splitting, syntax validation, and completion for Cloudberry.
  • Cloudberry-specific distributed and storage definitions are preserved and displayed where applicable, including distribution policies, partitioning, and table storage options.
  • Existing CloudDM workflows such as object management, authorization, auditing, and database CI/CD work for the supported Cloudberry scope, or unsupported capabilities are clearly documented.
  • Automated integration tests cover connection, metadata discovery, representative DDL/DML, and Cloudberry-specific table definitions.
  • Documentation states the supported Cloudberry versions, deployment/connection prerequisites, known limitations, and required privileges.
  • Existing PostgreSQL and Greenplum behavior remains unchanged.

Environment

Not applicable; this is a feature request.

Logs or screenshots

Not applicable.

Additional context

Official references:

Cloudberry supports PostgreSQL-compatible JDBC access and is compatible with PostgreSQL and Greenplum syntax, which should allow parts of the current PostgreSQL/Greenplum adapters to be reused. However, Cloudberry has its own product identity, release lifecycle, system catalog differences, and MPP-specific features, so a dedicated data source type and compatibility test matrix are preferable to silently mapping it to an existing type.

Checklist

  • I searched existing issues and did not find a duplicate.
  • I removed passwords, tokens, connection strings, and other sensitive data.

Metadata

Metadata

Assignees

Labels

No labels
No labels

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions