Power BI Best Practices for the Analyst/Developer (admin) in Enterprise SSAS Environment

[polymathy2017]

I've had difficulty finding specific information for working in an Enterprise SSAS environment. I have Fabric Admin rights, though that offers me very few privileges. I've been asked to help clean the dated and previously unmanaged Power BI environment utilizing current best practices for Power BI. I've worked in numerous Enterprise environments but this is the first time in an SSAS environment. Of course, we connect via a live connection so there's very little I can do other than create measures (though I can't validate them due data access restrictions) and view the model view in Power BI Desktop.

If I can't view existing measures from the datamodel, create virtual or disconnected tables, and don't have access to the data view (or any source data), is there any reason I should build my reports in desktop rather than in the service?

With that out of the way, looking for best practices in my current environment, so any help is appreciated. How can I get access to view measures and see the underlying data? Is it possible?

I'm responsible for determining which reports are being utilized but how can I do this without access to aggregated Power BI usage metrics? Can I backdoor it using either DAX Studio or Bravo?

First week in this environment and only received some data access late yesterday so still getting my feet under me. What tools do I need access to for providing senior-level analysis and gaining at least read access to the datamodel...SS Management Studio...Azure Data Studio...Visual Studio? The dev team uses Data Factory. Do I need that? 

In the current environment, considering data flows are happening in SSAS, is there any need to use Power BI's data flow functionality? 

Thanks in advance for any help provided.

[RossEdwards]

Based on what you've written.  You are doing the Power BI Reporting with a live connection to the Model in SSAS.  I prefer using Power BI Desktop for performance reasons. I found the cloud editor slower and frustrating.

If you wanted to look at the data model itself, get yourself Visual Studio Professional (you'll need SSDT as part of the installation and the Analysis Services extension) or Tabular Editor.  I found Visual Studio more user friendly but did work with someone who felt the reverse.  Depending on your permissions you should be able to download the model out and review it in either tool.  In my case, we use source control on our SSAS models and we get the models through the Solution Explorer in Visual studio.  Might be worth checking if your org uses TFS or Git for SSAS.

As for usage metrics.  Power BI offers you the ability to check the usage metrics of any report, going back about a month.  Power BI's APIs can also let you download usage out, which you could store locally and keep a history.

As for Power BI Dataflows, since all the data is already modelled in SSAS and you're just bolting a report on top, there is no reason to use Power BI Dataflows.

[polymathy2017 · Accepted solution]

This is awesome! Exactly what I needed. Thanks so much for the help. As I'm setting up the environment, or once it's setup, I may return with a question or two, if that's ok. I'll try to find the answers first before bothering you, though. 

Man, fantastic! Thanks again!