Earn the coveted Fabric Analytics Engineer certification. 100% off your exam for a limited time only!
Guys,
I am getting in the online service:
Last refresh failed: Sun Dec 02 2018 18:53:02 GMT+0100 (Central European Standard Time)
There was an error when processing the data in the dataset.Hide details
Data source error: | Not a legal OleAut date.. The exception was raised by the IDataReader interface. Table: NHTSA Complaints & Sales. |
Cluster URI: | WABI-EUROPE-NORTH-B-redirect.analysis.windows.net |
Activity ID: | b83dbb26-f4a8-4ef4-bb2e-f4c2357e0950 |
Request ID: | ecde4e07-5fea-6cd0-d13a-31a5e592a602 |
Time: | 2018-12-02 17:53:02Z |
My dates are completely fine. WTF?
No errors in PowerBI for Desktop. Why is it always so difficult to use Microsoft products?
DZ
Hi @danzrust
this error means that somewhere in the program is attempting to convert to or from an OLE Automation Date outside the valid range 1-January-4713 BC to 31-December-9999 AD. It might have slipped through because OLE Automation Dates are represented as a double.
Do you connect to data sources with Gateway and configure schedule refresh correctly?
If so, please check the actual date values in your table,
Is there any date value before 1/1/1900 or value like 0033, 0034 in one of the date fields in the table?
If it is, please correct that.
If all above don't help, please refer to some similar threads to get some ideas.
https://community.powerbi.com/t5/Integrations-with-Files-and/Not-a-legal-OleAut-date/td-p/9615
Best Reagrds
Maggie
I'd get a data dump of the most recent information between the 2 refresh times and have a good look at the dates. I know you have suggested it works fine in the desktop version. Some users just use the Preview Data in the Edit Queries and think everything is ok. Other users will do a full refresh and its possible there is a difference between the Desktop and Online mashup engines.
Without knowing the data, its pretty difficult to assess what is happening here, but if this has refreshed fine in the past, then the answer will lie within the new data.