Power BI dataflows (legacy) keep error : The credentials provided for the PowerBI source are invalid

[cst_bi]

Hi experts,

Pre-condition :

'Power BI dataflows (legacy)' setting with privacy level is 'public'

Cases :

Somehow in our dataflow can't show the table with error occured : 

The credentials provided for the PowerBI source are invalid. (Source at PowerBI.)

It happening with random table

When I 'configure connection' and click 'connect', nothing happened (error still exists)

But when I 'configure connection' and 'edit connection'

And then 'sign in'

 

The data table will show as below

But this scenario didn't last long because when I refreshed a few times, the table returned to its error state :

The credentials provided for the PowerBI source are invalid. (Source at PowerBI.)

These scenario above updated the 'Power BI dataflows (legacy)' setting to privacy level 'private'

So what is going on? 

Kindly please give advice to resolve it

Thank you

Sincerely,
Oviedityanto

[Anonymous]

Hi @cst_bi ,

This error usually occurs when the credentials stored in the Power BI service are not valid or have expired. To fix this, you need to update the credentials for the PowerBI source in the Power BI service. 

This should resolve the error and allow you to refresh your dataset. You can find more details on how to manage data source credentials here.

I have also found a similar post, please refer to it to see if it helps you.

Solved: The credentials provided for the PowerBI source ar... - Microsoft Fabric Community

How to Get Your Question Answered Quickly 

If it does not help, please provide more details.

Best Regards
Community Support Team _ Rongtie

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

[cst_bi]

Hi @Anonymous ,

This error occurs in dataflow. The dataset is normal

It's random error, please see the images below

1. as-is dataflow

dataflow : mps_source

table : _ppic_pemetaan_card_mps

error : The credentials provided for the PowerBI source are invalid. (Source at PowerBI.)

2. export mps_source into json file and import it to the same workspace

dataflow : mps_source(1)

table : _ppic_pemetaan_card_mps

error : -

This new dataflow "mps_source(1)" is still have no error free

The error "The credentials provided for the PowerBI source are invalid. (Source at PowerBI.)" occured on another table in this dataflow

In another example cases :

1. I have 3 table 'knsppk', 'oespb0' and 'wmsmakloon_transfer_log' in the same dataflow. The data display no error when open these table

But when i create new table using table combine of  'knsppk', 'oespb0' and then merge into 'wmsmakloon_transfer_log', the error occured at step 'Expanded wmsmakloon_transfer_log' as picture below

this is my step which error exists :

let
    Source = Table.Combine({knsppk0, oespb0}),
    #"Inserted text before delimiter" = Table.AddColumn(Source, "jenis_surat", each Text.Select([no_transaksi], {"A".."Z"}), type text),
    #"Merged queries" = Table.NestedJoin(#"Inserted text before delimiter", {"no_transaksi"}, wmsmakloon_transfer_log, {"c_no_transaksi"}, "wmsmakloon_transfer_log", JoinKind.LeftOuter),
    #"Expanded wmsmakloon_transfer_log" = Table.ExpandTableColumn(#"Merged queries", "wmsmakloon_transfer_log", {"dt_user"}, {"dt_user"})
in
    #"Expanded wmsmakloon_transfer_log"

My gateway connection for this dataflow is using 'Power BI dataflows (legacy)' as the picture below

In the picture above, my workspace is similar to 'module A' which is get the table source only from another workspace (workspace : data staging)

Sincerely,

Oviedityanto

[Ceeeg]

I'm having a very similar issue. I've got a dataflow loading tables from another dataflow, with a couple of computed tables built on top of the linked ones. Has been working fine for months until now. Basically all tables are getting the "The credentials provided for the PowerBI source are invalid. (Source at PowerBI.)" error. When I configure the connection, for most tables that seems to resolve the issue (although I need to do it individually for each table). However, a couple of the computed tables reference multiple other tables, and configuring the connection does nothing for them, even though the tables they refer to are already loaded into this dataflow and appear to be fine.

[cst_bi · Accepted solution]

Hi @Ceeeg ,

Are you using power bi legacy (no 1) connector? I've tried updated the connector using power bi dataflow / newer / not legacy (no 2) and the error dissapear

Here's my step to update connector from 'power bi dataflow (legacy)' to 'dataflow'

1. Open dataflow and remember the first table (in my case I choose 'wmsmakloon_transfer_log_natura')

2. Get data from 'dataflow' and select table 'wmsmakloon_transfer_log_natura' and load it. The table name will be 'wmsmakloon_transfer_log_natura (2)'

3. Go to 'advance editor' and copy these 5 line query from table 'wmsmakloon_transfer_log_natura (2)'

4. Now select 'wmsmakloon_transfer_log_natura' table, go to 'advance editor', paste and customise the code

This is my query before in table 'wmsmakloon_transfer_log_natura' :

let
Source = PowerBI.Dataflows(),
Workspace = Source{[workspaceId="19cfed95-xxxx-xxxx-xxxx-xxxxxxxxxxxx"]}[Data],
Dataflow = Workspace{[dataflowId="0100cd6a-xxxx-xxxx-xxxx-xxxxxxxxxxxx"]}[Data],
Entity = Dataflow{[entity="wmsmakloon_transfer_log_natura"]}[Data]
in
Entity

And this is my query after I customise based on table 'wmsmakloon_transfer_log_natura (2)' :

let
Source = PowerPlatform.Dataflows([]),
Navigation = Source{[Id = "Workspaces"]}[Data],
#"Navigation 1" = Navigation{[workspaceId = "19cfed95-xxxx-xxxx-xxxx-xxxxxxxxxxxx"]}[Data],
Dataflow = #"Navigation 1"{[dataflowId = "0100cd6a-xxxx-xxxx-xxxx-xxxxxxxxxxxx"]}[Data],
Entity = Dataflow{[entity="wmsmakloon_transfer_log_natura"]}[Data]
in
Entity

5. If you want to update another tables, copy these 5 line query from table 'wmsmakloon_transfer_log_natura' (which is customised table based on the new power bi dataflow connector). These query will be paste references for other tables

6. Paste it into another tables (picture below is example for table '

wmsmakloon_transfer_log_infion')

7. Delete unnecessary gateway 'Power BI dataflows (legacy)'

Hi @Anonymous ,

Please check this case, is it really because of the 'Power BI dataflows (legacy)' connector?

Sincerely,
Oviedityanto

[Ceeeg]

Hey @cst_bi ,

I've just been testing this out after reading your earlier response where you mentioned the legacy vs not-legacy connector.

Changing to the not-legacy connector appears to have resolved the issue.

To test it out, I exported the json, opened it in a text editor, and replace all instances of PowerBIDataflowsConnector with PowerPlatformDataflowsConnector. I then loaded that in as a new dataflow, and it seemed to make all the adjustments to the queries as appropriate. It asked me for credentials, I signed in, and it seemed happy. I opened up the manage connections window, and deleted the legacy connection. The privacy level on the new connection was set to private, and I couldn't seem to change that from manage connections, but then I went to the dataflow settings and was able to change it to organisational when I edited credentials there.

This is obviously not ideal, because loading it in as a new dataflow means it will have a different dataflow id, so anything referencing that dataflow will need to be updated. Luckily for me at least, the affected dataflow is the final one before it gets loaded for use, so I only have to update it in one or two reports. Seems like it would be very tedious if the problem had happened in a dataflow that is referenced by another dataflow and I had to update the dataflow id for each query in there.

Hopefully this doesn't all happen again when the Fabric Dataflows connector becomes the 'normal' one and the Power Platform version becomes 'legacy'.

[cst_bi]

update : 

We have removed the data gateway 'Power BI dataflows (legacy)' and create new one. The default privacy level is 'none'

Still the problem occured when edit the dataflow

The credentials provided for the PowerBI source are invalid. (Source at PowerBI.)

[cst_bi]

@GilbertQ , @ibarrau , @sulleyinoz , @aj1973 , @foodd , @collinq 

🙏