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

Register now to learn Fabric in free live sessions led by the best Microsoft experts. From Apr 16 to May 9, in English and Spanish.

App Dataset refresh fails with a table derived from a date column converted with locale

Originally posted here: https://community.powerbi.com/t5/Service/The-specified-culture-is-not-supported-on-an-extracted-date...

 

  1. Create a query table (A) with including a text column of dd/mm/yyyy values
  2. Create a second query list (B) of distinct values from the text-format date column from Table A.
  3. Publish (ok) and Refresh (ok)
  4. On Table A, on the appplicable column, Convert Type with Locale using English (Europe).
  5. Publish (ok) and Refresh (fail)
  6. Remove the Convert Type with Locale step on A, and add the equivalent on B
  7. Publish (ok) and Refresh (ok)
  8. A.column will be available online as a Text column, B.column will be available as a Date column.
Status: New
Comments
v-qiuyu-msft
Community Support

Hi @Mogura27,

 

in Step 4, do you mean change the local setting for Power BI desktop via via below option and convert the text type to date type for the column? 

 

q1.PNG

 

 

With this setting, I can refresh the dataset after publish to Power BI service. Please try to update your on-premise data gateway to the latest version then test again. 

 

Best Regards,
Qiuyun Yu 

Mogura27
Frequent Visitor

I tried updating the Gateway from 14.16.6549.2 to 14.16.6650.1 and the refresh still failed with the same error.

 

Underlying error code -2147467259 Table: Query2. // <-- this is the name of the derivative table

Underlying error message The specified culture is not supported.

DM_ErrorDetailNameCode_UnderlyingHResult -2147467259

Microsoft.Data.Mashup.ValueError.Reason DataFormat.Error

 

If the source table's query has the "Changed Type with Locale" step, the refresh will fail every time.

Mogura27
Frequent Visitor

This issue is still present with the May 2018 updates to Power BI. The only option I know of atm as a workaround for my pipeline is to do a stack of messy string chops and concats to flip the source data from DD/MM/YYYY HH:MM to MM/DD/YYYY HH:MM before converting the column into date/time data types to be able to chart on time axis/filters.