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

[tessahurr]

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

[jeroenterheerdt]

Hello @FraukjeC , thanks for checking in! @mamsteroonie is right, we changed the required permissions for Premium/PPU to just require Read permissions for viewing a report. Right now for Pro you still need Build. We are working on aligning that to Premium/PPU but it is taking longer than expected. 

[FraukjeC]

Hi @jeroenterheerdt and @mamsteroonie, thank you both for your information! We are indeed aware that within Premium/PPU licencing this issue is resolved. However, we have invested majorly in taking our full organisation to the next data-driven level by providing Power BI Pro licenses for all (5000+) employees only one year ago, so for us currenlty moving on to Premium/PPU licencing is not an option. 

@jeroenterheerdt I feel very optimistic hearing that Microsoft are actively working on removing the Build limitation for Pro users. This could mean that simply 'wait and work with the current limitation' might be a valid option for right now. Could you provide any time of timepath for this limitation to be fully dealt with for Pro users? Could we expect this within the current calendar year (2023) perhaps?

[jeroenterheerdt]

Unfortunately I cannot say any such thing in this public forum, but we are working on a blog post that should shed some light on what you're asking. (By the way, congrats on the 500th message in this thread!)

[FraukjeC]

It has come to our attention that this issue has been solved! I have not found any official Microsoft communication regarding this, but we discovered it ourselves and found a blog post by A Guy In A Cube comfirming this (start at 5:23): 
https://youtu.be/CIIzUvoUMOc?t=322

@jeroenterheerdt Would you happen to know if there is any documentation/release notes available somewhere on this?

[jeroenterheerdt]

this has indeed not been resolved for all workspaces / customers. We are still working on it. We expect more news in April.

[florenti]

I made a connexion with DirectQuery for Power BI datasets.

It's fantastic, very useful and powerful.

But, and it's a big BUT, once the connexion is made, there is no way to remove it!

Please, could you solve that ?

[jeroenterheerdt]

Hi @florenti thank you very much for you feedback and for trying this out! I am glad you like it! Regarding your question, I assume you have been looking to find the connection in the Power Query / Data Transformation window. It is by design that you will not find the connection there - as written in our documentation. However, you will be able to delete the connection using the Data Source connection dialog (Transform Data --> Data Source settings), in which you can also update the connection info if required.

Please let me know if this helps or if you are seeing anything else.

[florenti]

Thank you for this answer.
It works fine.

[bensycamore]

This has been in preview for 2 years now.  @tessahurr: What are we waiting for to make it GA?

[MiraNi]

Exactly my question! This was first published as a biggest thing in Power BI world at the time but now I hear nothing about this and not even a best guess for GA. It is annoying to know, that this feature is available but only in theory. There were so many bugs when testing this, that I am waiting for GA to continue any work with this.

[darrenfishell]

Five days from the two-year anniversary of "a milestone in business intelligence." 

I doubt that the mainstream media will recognize the importance of such innovation. But users that create and consume reports every day will immediately realize the impact of this change.

...

It is here, and it just works.

😫

[JoshAsh]

I would like to chime in by saying I support Microsoft in not making a feature GA until it is truly ready for prime time. There has been bugs yes but to their credit they have not released to GA prematurely and they are throwing fixes at the bugs quite quickly. 

Secondly, this is a very difficult thing to implement technically.

The feature looks to be largely implemented via extensions to the DAX language itself and implements relationships between the local and remote model by placing large lists of data inside DAX statements in dynamic sets.

Finally, if you are an enterprise model designer, it's important to understand how this feature works under the hood to help you decide whether or not you should use it in your particular scenario, taking into account its features and limitations - and other factors such as the size of your model. In some cases, extending a single, base model - is going to provide a far superior solution than composite modelling.

There is a setting DiscourageCompositeModels that can be used to control whether composite modelling is permitted at a dataset level.

https://learn.microsoft.com/en-us/dotnet/api/microsoft.analysisservices.tabular.model.discouragecomp... 

[jeroenterheerdt]

thanks all, we are not quite ready yet for prime time (or making this generally available). Yes, it takes a long time (longer than any of us want) but we are dependent on infrastructure changes that are taking longer than expected before we can declare GA.

[Domantas_G]

On the wesite it states that:
"If you refresh your data sources, and there are errors with conflicting field or table names, Power BI resolves the errors for you"

Not sure what is the exact definition of "field", but it doesn't resolve names of measures being idential. In my scenario I connect to multiple tables from different Power BI datasets (Direct query). Then if more than one of the sources gets new measure with the same name and they both hit my "combined model" at the same time, I get an error that "name is already used, and no refresh is working".

Feel free to contact me for a demo in case of any questions. 

[MrNobody]

I found the same when the same fully qualified DAX name for the measure in both models? i.e key measures[measure] is in both models. The refresh error implied it was getting duplicates when it was expecting a single item... I think the workaround I had was to subtlety change the table name.

But this does need to be resolved 

[jeroenterheerdt]

correct, this is on our radar and part of our de-duplication efforts.

[pedroamaro90]

Hi @jeroenterheerdt,

We are developing reports using the Direct Query connection for PBI dataset feature and now we are having the some use cases to them in Power BI Embedded, which is currently not supported as described on the documentation.

Is there any planned release to allows this kind of use cases? This is critical for us.

Thank you

[jeroenterheerdt]

what issues are you seeing?

[pedroamaro90]

This is the issue:

https://learn.microsoft.com/en-us/power-bi/connect-data/desktop-directquery-datasets-azure-analysis-...

Any plans to allow to use Power BI Embedded for reports with DirectQuery to PBI Datasets?

Thank you

[jeroenterheerdt]

There are no plans to enable this in the near future - it's a security problem that we still have to resolve. 

[than688]

Hi, this is a great feature - but after months developing for my org using the feature, I am suddenly getting errors when refreshing which is jeopardising my entire project:

I have only found one thread on this issue, here: https://community.powerbi.com/t5/Desktop/Refresh-Failed-file-pfoledbreader-cpp-line-836-function/td-...

with users positing that MS have removed the ability to create calculated tables on DQ data, or on using certain calculated table functions on DQ data (e.g. I am creating calculated tables on DQ tables using FILTER (with || OR conditions), CALCULATETABLE, and other DAX. All of this worked fine earlier this year, yet it now fails despite the model not changing.

Please could someone confirm what MS have changed to invalidate the methods I've been using - it's very frustrating and currently my only option is to convert all of these data models to 100% import, which will mean a huge amount of time wasted. Thanks!