Forum Discussion
Deployment pipeline based parameters for datasets reports and dataflows
We use the Deployment Pipeline feature to develop and test reports before moving them into production. A major challenge is that a change in a report may require a change in the dataset it is connected to, which in turn may require a change in the dataflow that we use in the dataset. Is there a way to set parameters/environment variables in a deployment pipeline and/or workspace that can be accessed from within reports, datasets and dataflows so that the data source connection settings in each will update dynamically when moving an object within a deployment pipeline.
For example, a report in the dev workspace should be connected to the dataset in dev; when we publish that report to prod in the deployment pipeline, we want it to automatically reset its dataset connection to the prod workspace dataset.
mbbozzuto, the connection between the items will happen automtically after deployment. Learn more about it here- Deployment pipelines, the Power BI Application lifecycle management (ALM) tool, process - Power BI | Microsoft Docs
Hi mbbozzuto
It could be that per changes in the M query, the parameters won't be recognized.
As suggested above, using the deployment pipeline's parameters rule is probably be the preferable way here.
5 Replies
mbbozzuto, the connection between the items will happen automtically after deployment. Learn more about it here- Deployment pipelines, the Power BI Application lifecycle management (ALM) tool, process - Power BI | Microsoft Docs
Nimrod_Shalit does a configuration need to be made to enable auto-binding or do the objects need to be in the same workspace but different pipelines (not sure if this is even possible)?
In our case, auto-binding doesn't appear to be happening automatically because we have reports in one pipeline and our datasets in another pipeline as described in the docs, and when we deploy the reports, they retain the connection to the dataset they were already connected to rather than auto-binding to the new dataset (i.e., report connected to TEST dataset, when report deployed to PROD, the report retains connection to TEST dataset rather than updating to PROD dataset).
Even if auto-binding can work, it would still be better to explicitly set a parameter variable to define the connection - in some cases, we need to use a variable, in other cases, we may want to explicitly set the connection directly to an dataset without a variable regardless of pipeline stage (eg. confirm that a report works with the PROD dataset while the report is in TEST stage before deploying to PROD stage).
Hi mbbozzuto
No configuration is needed. Just make sure that items are connected to eachother and both are on the same stage in different pipelines. for example:
Report on Dev stage in pipeline A is connected to dataset on Dev stage in pipeline B.
If that doesn't work for you so please open a support ticket since it's probably a bug we need to investigate.
As for adding parameter variables, it is possible to configure connections between dataset to dataset/dataflow using query parameters and then leverage parameters rules:
Deployment pipelines, the Power BI Application lifecycle management (ALM) tool, process - Power BI | Microsoft Docs
As for reports, please see here: Deployment pipelines, the Power BI Application lifecycle management (ALM) tool, process - Power BI | Microsoft Docs
BR,Lee
Hi mbbozzuto
It could be that per changes in the M query, the parameters won't be recognized.
As suggested above, using the deployment pipeline's parameters rule is probably be the preferable way here.
