Data structure for Reltio data sharing with Databricks - essentials
Learn more about the schema and streaming tables shared with your Databricks workspace.
Reltio data sharing with Databricks - essentials provides your Databricks workspace with read-only access to Reltio-managed streaming tables through Open Sharing.
The shared schema contains streaming tables for entities, relations, interactions, matches, merges, and links. You can access these tables from Databricks SQL, notebooks, dashboards, and AI and machine learning workloads.
Shared schema in your Databricks catalog
After the data share is received, a Databricks user creates a catalog from the share or mounts the share to an existing catalog. The user chooses the catalog name.
Within the catalog, the shared streaming tables are grouped in a schema named reltio_datashare_{datashareName}. The {datashareName} value is the name assigned to your data share in Reltio.
Reltio provides read-only access to every streaming table in the shared schema. These tables include type-specific landing and streaming tables for entities, relations, and interactions, as well as fixed-name landing and streaming tables for matches, merges, and links.
Data objects in the shared schema
The shared schema organizes each Reltio data type into its corresponding landing and streaming tables.
| Data objects | Landing table | Streaming table |
|---|---|---|
| Entity | entity_{entityType}_landingTable | entity_{entityType} |
| Relation | relation_{relationshipType}_landingTable | relation_{relationshipType} |
| Interaction | interaction_{interactionType}_landingTable | interaction_{interactionType} |
| Matches | matches_landingTable | matches |
| Merges | merges_landingTable | merges |
| Links | links_landingTable | links |
Data synchronization for each streaming table is triggered every four hours, with updates becoming available after the corresponding events have been processed. Landing tables are visible in the shared schema but are intended for internal processing and should not be used for downstream consumption or analysis.
Best practices for Reltio data sharing with Databricks - essentials explains which tables to query.