Lakshmi_M_Muthi's avatar
Lakshmi_M_Muthi
New Member
2 years ago
Status:
Needs Votes

Modify Deployment Pipeline

After Creating Power BI Deployment Pipeline, in future if we want to add a new workspace to the pipeline, we dont have ability.

Looks like we need to delete the whole deployment pipeline and recreate to add one more state.


It would be nice to have this functionality

5 Comments

  • If a deployment pipeline is built with 2 stages for instance and we need to add a third stage/workspace to the deployment pipeline, that is currently not possible. We shouldn't need to delete and recreate the deployment pipeline from scratch.
  • MaryanneBSA's avatar
    MaryanneBSA
    Frequent Visitor
    I discovered this the hard way. After configuring a deployment pipeline to promote from development to test, including deployment rules for multiple semantic models, to find it is not then possible to add a production environment to this existing pipeline. The only resolution currently is to delete the existing pipeline and all it's deployment rules, to then create a new pipeline with the additional environment and reconfigure all of the deployment rules again from memory (as you can't reference a deleted deployment pipeline and can't add the new one before deleting the old one). The consideration of creating all deployment pipelines with the maximum of 10 stages as a default position, allowing editing of workspace assignment at later date, has crossed my mind - but would only help if the stage is added to the end of the process. When using workspaces containing multiple lakehouses and semantic models, at best this requires very specific documentation of repointing per model and per stage, along with a ton of manual rework should a stage need to be added at the end or in the middle of the process (eg Training environment).
  • Absolutely a must have for deployment pipelines. In our organization we introduce new regions regularly for our customers to use our reporting solution and it would be nice if this can be achieved without re-creating a deployment pipeline. As a workaround we now created a deployment pipeline with as many stages as possible. But not an ideal situation.

Recent ideas