Looking up invalid data on a dense spool

[powerbi-dev]

I have a problem with a published dataset, with the error message "Looking up invalid data on a dense spool". It happens with a single measure and only in PBI Service.

The dataset was fine and this suddenly started to happen. When I'm using PBI Desktop, there are no errors, but I soon as I publish the data and connect my report via Live Connection, any visual with the measure returns the error. The measure is quite simple, just a SUMX with a simple CALCULATE filter, and it's used in many other measures of the report which do not return any error message.

I have already tried:

- Recreating the measure;

- Republishing the dataset;

- Deleting and republishing the dataset;

- Creating a new file with PBIT template and connecting/

None worked. 

If anyone has had this problem in the past, I'd appreciate some advice, thank you.

[Anonymous]

Hi, @powerbi-dev 

If you delete the measure and republish the report to Service , can the visual display normally ?

Best Regards,
Community Support Team _ Ailsa Tao

[Anonymous]

We are experiencing the same issue with several of our clients (different office365 environments). 

If we create the same report in desktop with import connection instead of Power BI dataset there is no issue. 
I have the idea it's an issue with Power BI datasets. 

[Anonymous]

We also have this issue with a report which is live connected to a Power BI composite model dataset. I have found that when I remove specific measures from the visual then it will load ok... problem is we need these measure in the report!

Power BI Service Error:

Power BI Desktop Error:

[Alexandre]

Same error for the last couple of days, on Canada Central cluster.

[MatherC]

I am having the same issue with a report in a premium workspace with data imported. The report was working in the service until this week.  It still works on the Desktop.

Measure causing issues uses trig functions within the MINX interator. I am dynamically calculating disance in miles between a selected zipcode and customer zipcodes using the Haversine formula.

https://stackoverflow.com/questions/27928/calculate-distance-between-two-latitude-longitude-points-h...

Activity ID: bb8615de-a753-400a-ba4e-177d423685b8
Request ID: 736f4cbe-7e92-4da6-bdb9-b74928aa5f28
Correlation ID: f4f0771d-45aa-d09e-e04e-6574769c1097
Time: Wed May 04 2022 15:20:21 GMT-0400 (Eastern Daylight Time)
Service version: 13.0.18173.56
Client version: 2204.4.11089-train
Cluster URI: https://wabi-us-north-central-c-primary-redirect.analysis.windows.net/

[spgarg04]

Same error, everything was working fine last week and still works on PowerBI desktop, however there is an issue with the PowerBI service

[spgarg04]

Activity ID: 7e815283-bdf9-42ed-9a45-42841d16b37c
Request ID: df5ab80e-0296-4ebc-b333-5cde64d98fb1
Correlation ID: 1c529c92-5332-66fd-bb3b-0cb383fdede2
Time: Wed May 04 2022 15:06:03 GMT-0400 (Eastern Daylight Time)
Service version: 13.0.18218.43
Client version: 2205.1.11147-train
Cluster URI: https://wabi-north-europe-j-primary-redirect.analysis.windows.net/

[iandrade]

We've been experiencing this issue for three days in a report from our workspaces. We've tried everything, but the problem persists.

[Anonymous]

I've had a call with support and they are aware of the issue and have provided an ETA of a fix of 5th June. Until then I've temporarily removed the measures from our reports. See below response from support:

As per verbatim we have understood that your are facing issue while refreshing the visuals in the report and you are facing the error as OLE DB or ODBC error: Internal error: looking up invalid data on a dense spool..
Currently few of our other users are facing the similar issue. An unexpected defect may have surfaced on our end, which caused issue, and our Product Group team confirmed that it is a known issue and they were actively working on this issue.
 
The team is already looking into the issue on higher priority. We will keep you posted with all the latest updates from them, we regret the inconvenience caused to you by this issue.
 
Please check the ETA for this issue.
 
ETA:  05/06/2022 

[ilyasss]

Hello all, 

I have the same issue.

I noticed that the RLS changes the number of visuals which cannot be displayed due to this error. 

When the RLS is on, I cannot see 7-9 visuals due to this error but when RLS is off only  2 visuals are not visible. 

Then when I removed the filters applied to those 2 visuals via the Filter pane the error is gone. 

So for my report, it seems that the problem comes from the filter context generated outside the measure (RLS / filter pane).