How to monitor the end user experience in Power BI Service?

[Toerstad]

Hi.

So we have created some great reports for our clients. They are managed, developed, tested, deployed and released  after all best practises known to mankind. Sources, infrastructure, network, pipelines, data stores and data models are monitored, incidents reported and handled.

Then - a call comes in. The user tells us a chart in a Power BI report is not working! (timed-out - or for some other reason)

We had made no changes, all our monitoring says everything is fine. ETL is executed with perfetion, data model prosessing too. 

Is there a way, a PBI Service monitoring API, or some other tool - that could have been utilized to monitor the end user's actual experience in the Power BI Service? A way for us to catch that - for some reason - this chart was not loading as expected? A way to probe and test if reports and dashboads in Power BI Service execute and render as expected?

Any advice would be greatly appiciated!

[lbendlin · Accepted solution]

Needs more details. Is this on Shared capacity or on Premium?

PowerBI-Tools-For-Capacities/Readme.md at master · microsoft/PowerBI-Tools-For-Capacities · GitHub

[Toerstad]

Hi, and thank you for your reply.

Yes, I understand this is low on details. In the most recent case I have been working on we do not have a BI Premium capacity - In other cases we do. This will, as I see it, be an issue any project of some size will have to solve in one way or another. My question was ment more as general question to maybye spur some discussion - rather than a support ticket, where I'm in need of a "perfect" solution.
I will however take a greater look at the git-repo you referenced - it looks very promising, and there may be something here we can untilize in some way or another. 

[lbendlin]

If you are running on Shared capacity there is no performance guarantee whatsoever.

From my experience it is more useful to work with your developers.  Reduce data model complexity, simplify Power Query queries, reduce query refresh duration, diagnose and optimize DAX queries, (drastically) reduce the number of visuals on a page etc.  This will very often lead to more agreeable end user experience.