Deployment pipeline: have multiple Prod stages for a unique pipeline

[Valeria]

Hello, 

I created a Deployment pipeline with Dev, Test and Prod. What I would need is different Production workspaces where deploying dashboards from the same Test WS. 

Allowing all PowerBi creators to publish on the same DEV stage and, after testing, choosing on which Production workspace I would like to assign a specific dashboard (as below).

Would it be possible?

Thanks a lot 

Valeria 

Deployment Pipline, staging 

[LeeBenjamin]

Hi @Valeria,
In case you're using 3 Prod environments just to seperate content to different audiences, you may find our Apps new feature as a proper solution. It is now possible to create an app to your single Prod environment workpace, and then assign each dashboard of this ws to different views (audiences) in your app. Learn more about this solution here:
Announcing Public Preview of Multiple Audiences for Power BI Apps | Microsoft Power BI Blog | Micros...

Best,

Lee

[Valeria]

Thanks @Nimrod_Shalit ! 

[Nimrod_Shalit]

@Valeria, we are looking into extending deployment pipelines to support more stages.

I don't have a timeline to share at the moment.

[jhiggs]

Any progress on this? trying to do the same as the original message, then we use an embed server to embed the reports into our app for our customers.

[LeeBenjamin]

hi @jhiggs,

We're working these days to allow creation of a customized pipeline. However, this enhancement will allow a creation of 2-10 serial (linear) stages. Split structure, like the "fork shape" in this thread, is not included. It is on our backlog but won't be released anytime soon. The customized pipeline we're working on will be launched in the upcoming months but I'm yet to have an ETA to share.

BR,

Lee

[Sonaldo]

"Split structure, like the "fork shape" in this thread, is not included"

Has there been any change or information regarding the schedule?

[LeeBenjamin]

Hi @Sonaldo ,
Unfortunately, we couldn't include it in this semester's scope, but it will be reconsidered around April.

[dim123dim123]

hi @LeeBenjamin 

Does this mean that 1 workspace can be assigned in several pipelines ?

[LeeBenjamin]

hi @dim123dim123,
It is not related to workspace assignment so no change there. Workspace-stage relation stays 1:1. no change.
The enhancement is for having the option to build a customized pipeline of several stages (2 to 10 stages) while today our deployment pipeline is of 3 built-in stages (dev-test-prod) that cannot be modified.

BR,

Lee

[Anonymous]

Adding my agreement here. Workspace per customer is a common pattern, needs to be a way to support that in pipelines. 

[dim123dim123]

@LeeBenjamin 

Sure, so it doesn't help us.
In our case we have 1 stage (DEV) and after that we need to advance it to many different TEST and PROD environments (client based ) .
So we are currently updating the BIM files through XMLA (it sucks). And now through the REST API I have to send PBIX files - first, of course, by changing the connection inside (in the testing phase).
In addition, we create process monitoring - data update, SQL stage checks, DAX automatic tests through XMLA 

[dim123dim123]

We have the same problem like @Valeria .

For us it is necessary to send from TEST to many PRODs by changing the database name in DATASET (DATASETS we can manage by XMLA and Workspaces by API ) .

It would also be good to be able to send directly to customers' Power BI when we have XMLA access to their workspace .

Right now automation pipiline process is very simple and not support real automatization cases

[mwroblewski]

Could you give more hints on that ? This is exactly the setup that we want to achive, having one dev/test and deploy to multiple prods by just changing the underlying datasets. How do you do that with XMLA / API ?

[dim123dim123]

Hi

 Unfortunately, at the moment I haven't found a good solution (we do semi-manually). So, after testing, we load the Model, after that we change the Datasource name and upload to multiply PROD over XMLA. Python script and pyadomd module. And we're looking forward to Microsoft and the Power BI TEAM doing something with PIPELINE