Issues in connecting to tabular model

[Wybie_TPG]

I'm encountering issues when I try to connect PBIDesktop(2.49.4831.521(x64)) or import data from our Tabular SSAS model (SQL server 2016 standard edition).

When I connect live to the model the project works fine for a little while, but after a few clicks it start giving the following error:

Error Message:
An unexpected exception occurred.

OData Error Message:
Failed to execute the DAX query.

When I want to import data it works fine for most measures and dimensions, but for some measures I get an error:

DataSource.Error: AnalysisServices: An unexpected exception occurred.
Details:
    DataSourceKind=AnalysisServices
    DataSourcePath=172.xxx.xxx.xx;CubeNameXX

The query that it tries to execute is = Cube.Transform(#"CubeNameXX",
        {
            {Cube.AddMeasureColumn, "Amount", "[Measures].[Amount]"}
        })

The srange thing is that we also use this measures in SSRS, without any issues, but after the PBI connection SSR Sdoesn't work either until we delete and recreate the measures. Than SSRS work fine, untill we again use PBI and the issue starts again.

I'm out of ideas how to solve this issue, does anyone encountered such issues befor or have ideas to solve it?

Many thanks!

[PDV_77]

I'm encountering the same problem with PBI. After clearing the cache (ssas) it works for i litle while. Then the same problems starts all over again.. Any ideas anyone?

[v-haibl-msft]

@Wybie_TPG @PDV_77

I cannot repro the same issue on my side by make live connection to my Tabular. Do you have a simple sample of Tabular SSAS model which can be used to repro?

Best Regards,
Herbert

[Wybie_TPG]

Hi Herbert @v-haibl-msft ,

Thanks for your reply.

 I isolated the issue to measures the are based on the following (DAX) expressions:

Amount:=round(sum('Journal Entries'[Amount_orig]),2)

Amount YTD:=TOTALYTD([Amount],Calender[Date])

Amount QTD:=TOTALQTD([Amount],Calender[Date])

Amount MTD:=TOTALMTD([Amount],Calender[Date])

Amount YTD Last Year:=CALCULATE([Amount YTD],SAMEPERIODLASTYEAR(Calender[Date]))

end so on..

On the other hand for example measures based on the following expressions on the same table are not giving any issues:

Day's Lag:=sum('Journal Entries'[Days_Lag_orig])

Journal Count:=DISTINCTCOUNT('Journal Entries'[DRank_JE_number_orig])

I also sended a Frown to Microsoft and they reply:

Thank you for the feedback. We are aware of the issue and have passed along the feedback to the team. Please keep an eye on Power BI Community for updates on this issue. If you need a follow-up with customer support, please file a support ticket atsupport.powerbi.com.

I still not understand wat is causing the issue, and since also the SSRS reports are infected it is hard to wait for an update.

I'm looking forward to your reply!

[PDV_77]

@v-haibl-msft it occurred (in my case) with measures sum(round(xyz)) en/or YTD functionality on the same measure based on a column with a large number of figures in numeric(18,2).  the strange thing about is that the error occurred only when I use PBI to visualize the measures (most of the time after a few times of clicking). After the error occurred the cache of ssas tabular seems corrupted (ssrs (and excel) starts at this moment also with throwing errors). Clearing cache or recreate/deploy the measures is the only temporally solution.

The same issue occurred if I connect PBI to my Workspace Environment (so deployment with wrong metadata could not be the problem). The problems are not just with the live connection feature but also occurred by importing the data in PBI.

Sincerely,

Peter

[v-haibl-msft]

@Wybie_TPG @PDV_77

I'm still not able to repro the same issue on my side. I created several YTD/QTD/MTD measures in my tabular. The Sep 2017 version of PBI Desktop always works well after I create and interact with the visuals.

@Wybie_TPG, have you got any other feedback from Microsoft? If not, you may need to get a SQL Profiler trace of the SSAS server when issue is repro'd and create a support ticket at http://support.powerbi.com (see bottom of page).

Best Regards,
Herbert

[Anonymous]

We have the same issue or atleast similar.

What I've noticed was that after a redeploy when everything is working as intended, by referring to the measure in a Power BI live connection embedded in a new measure, the error occurs again.

We have a simple SUM() measure that is not working when used Power BI Desktop during a live connection to a SSAS Tabular Model. It is however working if you browse the SSAS Tabular Model on the server and in a a Excel connection. It has happened that it stops working both in the model and in excel. Then a process full or redeploy has been necessary.

What is really strange is that if we recreate the same measure formula inside Power BI as a new measure, that measure work, but not the one from the model.

It seems that there is a problem with the live connection and tabular model that is throwing unexpected errors.

It would be really good if we could give some feedback on this. I have used profiler and the result is as follows:

*******

An unexpected exception occurred.

Query Text:
EVALUATE
ROW(
"TableName", 'TableName'[Measure]
)

*******

For the Measure that is created in Power BI the message is this:

EVALUATE
ROW(
"Measure", 'TableName'[Measure]
)

The Version we are using on the SSAS server is: 13.0.4206.0. It might be some CU-update that is required but havent found any information in the FIXes that solves this one.

Best regards
Gustav

[Wybie_TPG]

Hi Gustav, @Anonymous

Thanks for your reply! Unfortunatly we've not yet solved this issue either. As mentioned in my previous comments the PowerBI team replied that they are aware of this issue, but I did not yet receive details on what is causing the issue and when they plan to solve it.

Kind regards,

Wilbert

[Anonymous]

@Wybie_TPG

I will raise a ticket aswell regarding this. Will reply if we receive a solution.

Br

Gustav

[v-haibl-msft]

@Wybie_TPG

I can repro the same issue now. I’ve reported it internally to Power BI Team: CRI 48140880
I’ll post here once I get any update about it.

Best Regards,
Herbert