Hello,
let me introduce this post by explaining the setup. We have a pipeline set up with Dev, Test and Prod. In Dev we refresh the dataset with only a limited amount of data using some parameters. Once it is published to Test, we refresh the full scope of data to test, and once everything seems fine, we publish to Prod, where the report(s) are published to the users using an App. So far, so good.
Now here is my issue:
Every time we make a change to the report which affects the data model, the report inside the app "breaks" until the data model inside the Prod workspace is refreshed. In some cases, this can take over an hour in which users cannot really use the app/report(s). So far, I did not find a way around this issue, other than how we used to do it before deployment pipelines, which was making changes in PBI Desktop, publishing to a Test workspace, refreshing it in that workspace with full data, downloading the file from the service again, opening it on PBI Desktop again to eventually publish it to the Prod workspace. As you can read, this is super cumbersome with lots of manual steps... which is why I was super stoked when I discovered the pipeline feature - which, however, as described above, is quite the frustrating experience, too. Especially if you, for example, add a new field parameter to the report, which breaks entire visuals until the paramter is "initialized" in the respective workspace (or whatever it is called exaclty)...
So, IS there a better way of using the piplines to prevent the above mentioned from happening?
If there is not, it would be awesome if the service could at least be changed in a way that a change to the data model is not pushed to a workspace's app until the app is refreshed, as is already the case with changes to a report's "front end".
