Showing results for 
Search instead for 
Did you mean: 

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

344 REPLIES 344
Frequent Visitor

I just now wanted to try this out but cannot get it to work. There is one push dataset I connect to using the latest version of PBI Desktop (the preview flag is enabled). After adding the dataset, `Get data` is greyed-out and I do not have the option to add other data sources.

interesting, can you please open an issue so we can investigate?

New Member

Like the new Direct query feature and are using it extensively .. however Want to describe an issue we are having that is concerning for us as it makes testing changes to a dataset attached in direct query very difficult 


Normally we ..

step1 test modifications we make to a dataset by re-pointing existing ** reports at the modified ‘DEV’ dataset and Verify that they still work correctly .. if all good ..

step2  then we publish  overlay  ‘Prod’ dataset with dev dataset  .. and existing prod reports are good to go

**These reports are normally live connections to same dataset.


Issue is when the report is a  Direct Query local Model .. it throws an error in step1 (see below) even when the ‘Dev’ dataset is exactly

The same as prod  (so the ambiguous path error we receive is not valid error)… if we go ahead and overlay prod with dev … the prod reports

don’t throw an error and work fine .. but if they don’t I’m afraid we have no way to recover . as we can’t connect to an alternate datasource.  In fact I suspect that even if we restored the prior Version of the dataset the direct query local model reports would not be able to connect to it.. and we would have to re-create them basically from scratch


Pls advise if you have heard of this issue and if there is a workaround..


**Note Datasource that we are replacing in the direct query local dataset  is an incremental



Error Returned 




so, the error says you have an ambiguous path. Is that true? Is so, you will need to redesign it.

no ambiguous path .. it is an exact copy of the dataset being used in the direct query model published under a different name in a different workspace.  

sorry for the late reply - please open an support ticket so we can investigate this.

Frequent Visitor

Hi team,

we had the following scenario:
We created a "golden dataset" that we use in some dashboards. If we now connect an additional data source to the dashboard an extra dataset is created for it, which makes sense, yet we have a strange phenomenon. We have set the permission for both datasets to "Read". But the users can't see the data of the "Golden Dataset" in the new dashboard. Only when we set the permission of the users in the Golden Dataset to "Build", they can see the data. Of course, this is not desired because we want to prevent users from creating their own dashboards with the dataset.


Are you aware of the problem ?




this is a known issue and we're investigating a fix.

Frequent Visitor

This is a known issue, please upvote this idea to help prioritize it:

Microsoft Idea  · Directquery for datasets only needs read permissions on all datasets (


Frequent Visitor

Hi MLechner, 


We have the same behaviour and to my knowledge there isn't another way around it at the moment, which is frustrating...

Regular Visitor

This is a great feature, when using/testing it I run into the following problem though.


- I connect to an Azure Analysis Services tabular model

- Then I add a calculated column to a fact table using RELATED to add a value from a dimension table within the same model which has a one to many relationship with the fact table, no errors appear when doing this

- As soon as I try to use the calculated column (or something using it) in a visual this error appears:




- The strange thing is that this does work when doing the same thing but with a different dimension from the model (no error appears when using it)


What can cause this error? As far as I can tell I'm not doing something that isn't supported and as I'm referring to a table within the same model I don't see any issues (no limited relationship).


One thing that is different between the dimension for which this works and the one for which it doesn't: the key in the fact table which relates to the dimension which doesn't work has blank values as the dimension isn't applicable for all the rows. In this case I expect RELATED to return a blank value. Also I'm pretty sure this worked when I tested it earlier in the year.


Hopefully you can help me with this, thanks in advance!

we were unable to reproduce this - can you please open an issue for this so we can collect some more info?

Sure, do you mean I should post it on Issues - Microsoft Power BI Community or should I open a support ticket?

either works - but support ticket is faster.

hmm, interesting, let me talk to the engineers. Thanks for reporting!

Frequent Visitor

This is mega!

Do we have a view when the next version of SQL Server will be released brining this feature online for tabular SSAS users?

Working on the Product Team for PowerBI must be the coolest job in Microsoft.

it is pretty cool, thanks 🙂 regarding the next version of SQL - not sure, but looking at the speed of previous releases I'd say 2022/2023.

New Member

To be blunt, the lack of conformed dimension support, the errors, the bugs, the poor performance on all but tiny datasets... This feature is useless to enterprise customers. 


Please prioritise quality over speed to market.

Josh Ash


thanks for the feedback - it would be great if you can give specific examples of where you are seeing perf issues, errors, bugs and on which version of Desktop. We have worked on improving performance specifically in the last couple releases of Desktop and our telemetry shows that those improvements are paying off. Curious to learn what you think we need to do as well. Also, not sure what you mean that we don't support conformed dimensions. We are working on a best practices document that explains how to deal with the most common issues people have with this feature.

Advocate I
Advocate I

Has there been any further thoughts around the end user permissions that are required in order to view a published report that has been crafted from a shared dataset under the Direct Query / Local Model? 


Having to add a report end user / consumer to the primary shared dataset with Build permissions is an issue. We have analysts that we restrict to shared datasets and in turn they share out their reports to end users. Having the analyst grant users access to the report they created and then turning to the admins to have the end users granted Build permission to the shared dataset is a security issue. These end users could then connect to the core shared dataset and access data they should not be able to view.  

Helpful resources

PBI_User Group Leader_768x460.jpg

Manage your user group events

Check out the News & Announcements to learn more.

MBAS on Demand

2021 Release Wave 2 Plan

Power Platform release plan for the 2021 release wave 2 describes all new features releasing from October 2021 through March 2022.

Get Ready for Power BI Dev Camp

Microsoft named a Leader in The Forrester Wave

Microsoft received the highest score of any vendor in both the strategy and current offering categories.

R2 (Green) 768 x 460px.png

Microsoft Dynamics 365 & Power Platform User Professionals

DynamicsCon is a FREE, 4 half-day virtual learning experience for 11,000+ Microsoft Business Application users and professionals.

Top Solution Authors
Top Kudoed Authors