Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 

Earn the coveted Fabric Analytics Engineer certification. 100% off your exam for a limited time only!

Reply
tessahurr
Employee
Employee

Share your thoughts on DirectQuery for Power BI datasets and Azure Analysis Services (preview)

Hit Reply and let us know what you think of the DirectQuery for Power BI datasets and Azure Analysis Services.  To learn more about this feature, please visit this blog post or our documentation.

 

Here are some areas that we'd like to hear about in particular:

  • Performance
  • Query editor experience--the remote model query doesn't show up in the query editor and only in the data source settings dialog. What are your thoughts?
  • Navigator experience
  • Thoughts around governance and permissions for models that leverage this feature
  • Nesting models, i.e. building a composite model on top of a composite model
  • Automatic page refresh for live connect in composite models

Thanks and we look forward to hearing your feedback!

 

- The Power BI Modeling Team

531 REPLIES 531

can you please open a support case for this? If you can't let me know and we figure something out.

Hi @jeroenterheerdt@MBonnett@jane01,

 

was the issue resolved with helping of the support ticket? We still have the same issue in our org.

 

Thank you for an update.

Hi @soldous - yes they did get to the bottom of it! For our org this was fixed by enabling the tenant setting (we only did this for specific users): Allow XMLA endpoints and Analyze in Excel with on-premises datasets 

frostystoo
Frequent Visitor

This is the feature we are been waiting for for literally years, thank you! However, it doesn't seem to work, does something need to be enabled on our tenant?

 

Specifically I have enabled the preview feature, restarted Power BI desktop and connected to one of my Power BI datasets (in fact a few of them) but all the options to import data are greyed out as it always used to be. I also don't see the "make changes to this model" link in the bottom right as described in the video. A colleague of mine has also tried but the same thing.

@frostystoo Not all models are supported (for example SSAS, even if published to the PBI Service). Can you make sure your dataset is supported by checking the limitations section here: DirectQuery for Power BI datasets and Azure Analysis Services (preview) | Microsoft Power BI Blog | ...

If you believe it should be, please share some additional details/screenshots to help us identify the problem.

Mariusz
Community Champion
Community Champion

Hi @tessahurr 

 

First of all thanks for this amazing future, I have seen the demo ( during MBAS 2020... I think ) and was waiting for the preview with anticipation.

I do have a few questions I hope you will be able to answer: 

While connecting to two datasets Sales and Order ( both star schemas ) and both datasets/models contains the same dimension Product: 

  • How do I remove one of these dimensions?
  • Will it make a difference in performance if I build relationships between two different datasets ( Product from Sales dataset  filtering Orders fact from orders dataset )
  • Will you be able to merge dimensions if identical?
  • Will this stay as Pro or will this be a premium only future when generally available?

image.png

Many Thanks 

Mariusz

1. Can you give more info on the data sources you are connected to for Product and Product 2?

2. I don't believe so, but perhaps @jeroenterheerdt can help answer your question about performance.

3. This is an interesting idea. We don't allowe for that right now, but can you describe a bit more about your expected behavior?

4. When this feature is GA it will be available for both shared and premium.

 

Thanks!

I have the same question as Mariusz, and I think it gets at architectural issues that Power BI has with conformed dimensions. It's never been clear to me the best way to handle them as there is no concept of shared dimensions like there used to be in SSAS MD.

 

If there's two models, both with a copy of the same Product dimension and those two models are "composited" together, you're going to have two Product dimensions. How do you get the two models to work together across this dimension? Can you remove one? Can you link them together somehow? It seems to me that each model may contain its own measures with references to its own Product dimension. How do we keep these working?

conforming dimensions is something we are still working on.

Well, is this realy related to "Composite"? Same question with "Import"...
The usual way is to have just one dim with several relations and to build measures with "USERALTIONSHIP".

Hi Oliver,

 

I'm not sure I follow you. I'm talking about multiple models, not multiple tables within a single model.

 

Model A has a star schema that includes Product dimension.
Model B has a star schema that includes Product dimension.

Model C is a composite of A and B. C now contains Product(A) and Product(B). What's the right way of handling this situation?

@tessahurr I am not sure who @jeroent is, but that's not me 🙂 regarding the performance question from @Mariusz: if I understand it correctly you are asking if there is performance difference between filtering within one source compared to across two sources. While the answer is it depends, I would expect the performance for filtering across two sources to be less than within the source, unless one of the tables is significantly larger than the other.

MCornish
Responsive Resident
Responsive Resident

I get this when converting to a local data model

 

Capture.PNG

BCRio
Frequent Visitor

This is a brilliant enhancement with huge potential value - thank you.

 

Note however, preview testing I have an issue in desktop importing a PBI dataset into a simple existing local model  (i.e. composite model) where the table relationships from this "live connection" dataset were lost when using this method importing as direct query.   The imported PBI dataset tables are all shown in the Model view and data can be obtained from each individually - however these tables have no relationships and neither can the original relationships be re-added, so the dataset model cannot really be used.

I'm not sure this is due to the nature of the preview version, a current table limitation or something I am doing incorrectly  - does anyone have any suggestions?

Anonymous
Not applicable

Great announcement! Thanks Power BI team!

 

LIMITATIONS

1 - I was testing how many online dataset I was able to add to a single model. Seems that 2 is maximum right now. When trying to add a 3rd, I'm getting the following error message: "Sequence contains more than one element".

 

UX/UI

2 - It would be really nice to have the tables with the same color the model is. If it's crowded it will be nice to quickly and visually identify or match model with corresponding tables.

 

pbi error.png

Anonymous
Not applicable

Also to add seems like "RELATED" formula dont work. is that normal?

Anonymous
Not applicable

Im getting issue publishing the report.

"An error occurred while attempting to publish xxxx
Publishing failed for an unkown reason.
(PowerBINotAuthorizedException)"

cyrusemami_0-1611584456824.png

I get the same error message with a Power BI dataset whose storage mode I changed from Live to Direct Query (Mixed).
In Live mode, the upload works normally - just like with other datasets (Import, Live, ...).

 

 

are you sure you are logged in?

Anonymous
Not applicable

Yes Im sure i logged in.
I tried it yesterday and it works. today im getting this error.

Are you sure you did not change anything between today and yesterday? Different dataset? Different workspace?

Helpful resources

Announcements
April AMA free

Microsoft Fabric AMA Livestream

Join us Tuesday, April 09, 9:00 – 10:00 AM PST for a live, expert-led Q&A session on all things Microsoft Fabric!

March Fabric Community Update

Fabric Community Update - March 2024

Find out what's new and trending in the Fabric Community.