Realtime Compute for Apache Flink uses catalogs for metadata management. Catalogs register, store, and load metadata from various data sources, enabling unified access to external system metadata through Flink SQL.
Why use a catalog
Unified management: Centrally manage metadata from different systems to simplify data discovery.
Simplified operations: Reference table functions directly in SQL without repeated definitions.
Improved efficiency: Reduce metadata hard coding to increase development efficiency.
Integration: Seamlessly integrate with external metadata services such as Paimon and Hive.
Supported catalogs
Traditional data lakes/warehouses | Relational/analytical databases |
Message-based/non-relational data sources | Integration / general extensions |