Linda Sarace on 05 Nov 2019 04:30:30
Our policy is perform dev and test efforts separately from production, therefore we develop and test data sets and Data flows in test work spaces. Once we are satisfied with their behavior and performance, we publish them into the production workspace. With PBIT files, it is easy for us to publish approved data models and reports into the production work spaces. Because data flows are built within PBI Service and have no file association, it is difficult to migrate them from test to production. We currently have to rebuild the entire Data flow in production, which is risky an cumbersome.
- Comments (1)
RE: Better deployment method for Dataflows
Can you not just export the JSON definition of a dev/test Dataflow and create a new Dataflow in the production environment by importing that JSON definition file? I appreciate that you may still need to make a few small tweaks by hand for things like data sources (if development/testing environment includes its own isolated data source to avoid exposing sensitive data to the wider dev/test team) but beyond that, what are you having to rebuild?