Deployment pipelines - multiple pipelines working together


This content is archived and is not being updated. For the latest documentation, go to What's new in Power BI?. For the latest release plans, go to Dynamics 365 and Microsoft Power Platform release plans.

Enabled for Public preview General availability
Admins, makers, marketers, or analysts, automatically Jan 19, 2022 -

Business value

Many BI teams separate data items (datasets, dataflows) from visualization items (reports, dashboards). These are managed in different workspaces and sometimes by completely different teams, with different release processes. However, since these items are related by data connections, BI teams need an easy way to manage and connect them across the different lifecycle environments.

This feature enables different BI teams to manage items that are connected to each other in different pipelines, while also ensuring that those items will be connected across their respective environments.

Feature details

Today, object connections are persisted in a pipeline. What this means is that whenever a report and dataset are connected in Test stage, they'll be remain connected to each other when deployed to Prod stage.

Going forward, we're extending this capability across different deployment pipelines. Here are some examples:

  • A dataset and a report are managed in two different pipelines and are connected in Dev stage. When that dataset and report are being deployed (each one in its own pipeline) to Test stage, Power BI will maintain the connection between the dataset and the report.
  • If you try to deploy the report into Test stage while the dataset in the second pipeline doesn't yet exist in Test stage (wasn't deployed), the deployment will fail. This is the same behavior we see in single pipeline deployments today.
  • If a report points to a dataset in a different pipeline and in a different stage (for example, Prod stage), then after deploying the report to Test stage, the report will retain its connection to the dataset that's still in Prod stage. This allows BI teams who build reports only on Prod data to continue working with pipelines.

The examples above are for reports and shared datasets, but this feature will apply to any combination of items connected to each other and supported in deployment pipelines:

  • RDL reports--> datasets
  • Datasets--> datasets
  • Datasets--> dataflows

See also

Feature exploration (video)