I'm looking for some advice on the best practice for making changes to an MDM data model. We have multiple environments (dev, test, prod). Some changes to the model will truncate the existing records even if the entities are basically the same, so the data needs to be reloaded. I could build a DI job to move data from test/prod to the new model in dev, but is there a simpler way; using the data containers or something else? When the final changes are deployed in production, the production tables will be truncated. Where does the backup data come from?
Any thoughts, advice or links to relevant information would be appreciated.
The Studio Data container can export data records for entities residing in MDM server. (Right click data container name under "Data Container" -> Export content from MDM Server). You can then import (Right click data container name -> Import content to MDM Server) the records to the new environment (you should have a corresponding server location defined in studio), you can pick the entities that data structure have not been changed. Please be aware that if the entities you are importing have links to other Entities (Foreign Keys), they have to be imported together, otherwise, there database will complain about the FK constraint not satisfied.
Hope it helps.
Thanks. That solution works fine, given there are no data model changes and you are trying to migrate fewer than 10,000 records. The question is more about how to move a breaking model change through the dev/qa/prod landscape without losing data. What is the best practice for this?