Query/Question:
Why does Databricks Datashare configured with ovOnly=true output only a single scalar value instead of an array for multi-occurring nested reference attributes?
Probable Cause:
In ovOnly=true mode, Datashare generates a flattened schema. Because the data model does not contain a separate multi-occurring indicator for nested attributes, the pipeline relies on the configured survivorship strategy: non-aggregate rules (such as LUD) flatten attributes to a single scalar winner, while Aggregation rules preserve them as arrays.
Recommended fix:
Update the L3 configuration for the relationship/nested attribute group to use the Aggregation survivorship strategy, then re-sync the entity/relationship types to the pipeline.
Documentation links:
[Datashare Schema Flattening & ovOnly Mode] (https://docs.reltio.com/en/explore/reltio-data-pipeline/datashare-flattened-schema) | [Configuring Aggregation Survivorship] (https://docs.reltio.com/en/model-data/survivorship-rules/aggregation-survivorship)
Comments
Please sign in to leave a comment.