The ultimate Fabric, Power BI, SQL, and AI community-led learning event. Save €200 with code FABCOMM.
Get registeredCompete to become Power BI Data Viz World Champion! First round ends August 18th. Get started.
So, Ive created a dataflow which queries a SQL database via a view, the dataflow has a scheduled refresh and, although the query is not exactly graceful and the return is pretty large, this seems to operate fine
I then have a dataset connected to this dataflow (and others). This dataset merges the data from the seperate dataflows together into a singular table, again with a scheduled refresh setup. However, every time its refreshed, after a lengthy, but not suprising, time (it only needs refreshing once a day) I get the following error.
The credentials provided for the PowerBI source are invalid. (Source at PowerBI.) Table: [Table name in the dataset]
Strangely if I modify the SQL view-code to limit the rows returned by the view, without changing anything in the dataset, then manually refresh the dataflow with the now reduced set of data, followed by a manual refresh of the problematic dataflow, it will refresh without issue. NB this is without changing anything in the dataset, such as the schema or republishing it, this is only the number of rows handled by the service
If I increase the view back up to the original size, refesh the dataflow and the dataset refresh fails again
Just to be thorough, I have checked credentials in the gateway and also the ones used to connect to the dataflow and all are fine. Though -this would seem unneccessary as the credentials are proven fine by it working with the slimmed down view
I could understand this issue, if it was related to the size of the data processed by the dataset or dataflow- but I find the error relating to credentials confusing
Thanks
HI @Anonymous ,
AFAIK, if dataflow not created in premium workspace, it will has 2 hours refresh timeout, please confirm if refresh time is exceed to the limit.
Regards,
Xiaoxin Sheng
Hi @Anonymous thanks for your reply
The dataflow refreshes in about 10 minutes
Regardless, it is the dataset erroring - not the dataflow - and that takes around an hour.
Thanks