Power BI Life cycle Management

[aschattopadhyay]

Hi All:

I am pretty new to Power BI and I have been assigned the responsibility of setting up some governance around Power BI in our organisation. We have already delivered some POC with reports, dashboard etc and users have really loved them. Now it is the time for the next phase. 

We have got one Power BI tenant signed up for few users and we have got PRO licenses for the users. Just wondering whether I can get any help / suggestion in terms of setting up some best practices to manage life cycle of Power BI projects.

By that I mean what's the best way to achieve the rollout of Power BI applications in terms of DEVELOPMENT, TESTING, UAT and PRODUCTION environments. As we have got only one environment, we need to go through this normal SDLC process of DEV, TEST, UAT and finally roll out the changes to PRODUCTION. Is there any way to achieve it in a Power BI tenant? Else It is becoming very difficult to manage. Can we have things like multiple workspaces for DEV, TEST, UAT and then share them among the developers and finally push the changes/dashboards to PRODUCTION workspace? Just an idea.

Any guidance will be much appreciated.

Thanks and regards,

Anindya

[sherrynaeem2]

i've had some really interesting challenge to this!

In my scenario. I have developed a power bi report using OData feed so it uses a '.svc' link. This report has been uploaded to 3 workspaces namely dev, test and uat. The data is accessed through OData feed which is on premise and accessed via gateway.

Question is how do i get my report in a worksapce to point to correct datasource?. OData link is also same in all enviroments therefore datasource defined in gateway is also same. 

[andrewporter]

As a manager/ technical lead, I view Power BI as a fresh paint-job on Excel (PowerPivot, PowerQuery, PowerView) basically hosted in somethig like SharePoint. You may find that you can ease up on SDLC in favor of some type of more informal "Report Certification" process led and validated by your Power BI Champion.

Microsoft provided a pretty extensive white paper on Power BI governance and lifecycle approaches. 

Business-Led Self-Service BI - Business supports all elements of solution (preparation, modeling, report creation and execution)

IT-Managed Self-Service BI - IT owns data preparation, modeling. Business owns report creation and execution.

Corporate BI - IT owns all aspects of preparation, modeling, report creation. Business just executes reports.

https://powerbi.microsoft.com/en-us/documentation/powerbi-admin-governance/ 

They go very in depth on roadmaps, and the pros and cons of each approach. The most important factor across all approaches is identifying the Power BI Champion. They are likely technical or quasi-technical, understand and define process / vision for Power BI, and are of course passionate about building a data-driven culture.

I believe that this post may sway towards the Corporate BI approach, and I would encourage you to think outside the box, towards a more business-led approach. Pie in the sky, but in my mind it is important to encourage teams to build their own reports at the very minimum, and strive for a deeper understanding of their company's unique data. On multiple occassions, I saw the light-bulb go off when data modeling, slicing and dicing, and effective quantitative visualization come together.

Power BI shouldn't just be a flashy upgrade from SQL Server Reporting Services. It's a different paradigm that belongs more in the hands of the business. It has the potential to build let IT focus on more impactful projects and governance rather than report development.

[Anonymous]

too bad that link https://powerbi.microsoft.com/en-us/documentation/powerbi-admin-governance/%C2%A0 doesn't work anymore.  where should it point to now?

[jagatho · Accepted solution]

This blog entry describes a way to achieve a DEV, TEST and PROD environment in power bi as well: http://angryanalyticsblog.azurewebsites.net/index.php/2016/04/17/power-bi-content-workflow/

So in a corporate Power BI environment, I currently think you basically have the following options: Let you end users access reports and dashboards through content packs or throught a PROD group workspace (with privacy set to members can only view Power BI content).

With content packs you can include just the content that fits your user profile, but if they for example need access to a lot of reports from multiple business areas (ie. TEST workspaces) it gets messy in their personal workspace, since the separation is not visible in their personal worspace. Another difference is that with content packs, the end user can create a copy of the reports and modify it.

[aschattopadhyay]

Fantastic link! Makes perfect sense.

Thanks a lot 🙂

[aschattopadhyay]

Hi Konstantinos/Ankit :

Thanks for your help and information. It was definitely good to come across those information.

So I think as best practice, we can have 2 models :

1. In a Corporate BI environment, multiple environments typically exist for managing the lifecycle of BI assets:

   • Development: Available to IT BI developers only
   • Quality Assurance (QA): Available to a select group of users for purposes of user acceptance testing
   • Production: Available to relevant functional users based on job role, subject area, department, project, etc.

      Typically separate Power BI group workspaces are utilized for Dev, QA, and Prod purposes.

    2. IT-Managed Self-Service BI: This is where IT will be responsibile for providing the data. However the data managed by IT         will enable Self-Service reporting. In this case we will have no Dev, QA and Prod cycle.

What do you guys think? I want to continue this discussion.

Regards,

Anindya

[aschattopadhyay]

Any further comment on this one?

Would love to see how people deal with the scenario in real life and what sort of governance/process they have set up.

Thanks and regards,

Anindya

[ankitpatira]

@aschattopadhyay On top of what @konstantinos said, you should consider using power bi templates and query parameters features. They were introduced in recent power bi desktop update and are meant to provide control to users to manage power bi reports within their organisation. 

[konstantinos]

You can check this whitepaper that describes governance in PowerBI..This is the most complete so far

https://powerbi.microsoft.com/en-us/documentation/powerbi-admin-governance/