Unable to connect to Data Lake Gen 2 based Databricks table from Power BI Desktop

[Anonymous]

Hi,

I have two tables defined within my Databricks Cluster on Azure.

1. Table 1 is pointing to local file storage.

2. Table 2 is pointing to a Azure Data Lake Gen 2 storage. This storage is mounted using persistent configuration. 

Within Power BI Desktop, I could successfully connect and Direct Query to Table 1 however I get error while connecting to Table 2.  

DataSource.Error: ODBC: ERROR [HY000] [Microsoft][Hardy] (35) Error from server: error code: '0' error message: 'com.google.common.util.concurrent.UncheckedExecutionException: com.databricks.backend.daemon.data.common.InvalidMountException: Error while using path /mnt/xxx for resolving path '/yyy.csv' within mount at '/mnt/xxx'.'.
Details:
DataSourceKind=Spark
DataSourcePath=https://....
OdbcErrors=Table

I am basically attempting a setup where Power BI will Direct Query to Databricks Cluster which in turn will access Azure Data Lake Gen 2 Storage.

Any advise on resolution?

Thanks,

S

[andre]

have you been able to figure this out?  i am running into the same issue

[v-yulgu-msft]

Hi @Anonymous ,

Please refer to this document:

Power BI Desktop customers cannot access dataflows stored in an Azure Data Lake Storage Account unless they are the owner of the dataflow, or have been authorized to the CDM folder in the lake.

Regards,

Yuliana Gu

[Anonymous]

Hi,

My scenario does not use Power BI dataflow to connect to DL Gen 2 directly.

I am working on scenario to connect to Databricks table from Power BI. I could connect to natively hosted table ok however table hosted on \mnt (in my case, DL Gen2 location) is throwing error.  

Thanks

S

[alfred1]

me too, got the same error and the same scenario

[alfred1]

@Anonymous  did you find any solutions or answers?

[Anonymous]

Not yet. I am not 100% sure but it looks like connection to "tables that use mounted storage" is not yet available. 

[andre]

ok, the solution is to make a change to your cluster configuration...

details are here

https://businessintelligist.com/2019/04/11/how-to-make-azure-databricks-work-with-azure-data-lake-st...