Skip to main content
The Cognite EDM extractor translates EDM data entities to store and maintain in CDF. Data is retrieved using the EDM native data model, and the extracted objects will match the EDM OData entity schema. This translation is automatically done inside the Cognite EDM extractor. There are two EDM entities categories:
  • Incremental entities have a timestamp field used by the extractor to handle and control incremental runs.
  • Non-incremental entities don’t have a timestamp field. These entities are always updated and inserted in full load into CDF.
When extracting an entity, the extractor also retrieves any other entities it relates to that aren’t in the incremental category (in a transactional manner). This means that the extracted entity written to RAW contains the full context of the entity, avoiding the need to join multiple RAW tables. The Number of related entities column in the list of entities specifies how many related entities an entity references. This can be overridden on an entity-by-entity basis (such as to reduce the load on EDM) by either turning off getting related entities completely or specifying which related entities to include in the configuration through the includeFields or excludeFields. The Cognite EDM extractor supports these EDM entities:

​
Incremental entities

​
Non-incremental entities

These entities don’t have a create/update timestamp field, and an extractor can’t detect changes without downloading all the content.
Last modified on December 9, 2025