Connection to Azure SQL Server give oledb errors

[Anonymous]

Hi All,

When I try to select to several tables in Azure SQL Server from Power BI Desktop on premise and then it trys to connect I get oledb errors and it does not connect but if I try to connect one at a time no errors and it can connect. Problem is to connect to each table requires to renter my server name each time since I go through getData.

Anyone see anything like this before?

Thanks!

[v-xuding-msft]

Hi @Anonymous ,

What mode do you use to get data? Import or Direct Query? I can't reproduce your issue with the two modes. Can you please share the entire error message?

You could try to change the credentials and privacy level (to None or Organizational) firstly to check if it works.

Best Regards,

Xue Ding

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[Anonymous]

Hi,

It was import. I tried again today and it did not repeat but the one difference is the credentials were aready stored in Power BI from the previous import for a different report. Maybe that is the difference? How do I delete the credentials before I do a get data?

Thanks,

-Tony

[Anonymous]

Hi,

I followed your image above to delete the credentials and it did not clear the dialog box of the existing credentials. I stopped then restarted Power BI and the credentials were gone. The I repeated get data and it asked for credentials and loaded the tables into Power BI and no errors this time.

Best Regards,

-Tony

[v-xuding-msft · Accepted solution]

Hi @Anonymous ,

Glad that your report works fine now. For my image, it shows how to edit permissions. If you want to delete, you could try the option of "Clear Permissions". 

Please mark the helpful answer as a solution if you feel that makes sense.  Then we will be able to close the thread. More people will benefit from here.

Best Regards,

Xue Ding

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.