Spark Hdinsight and DirectQuery bug with null values

[clackner]

I'm trying to connect to data tables on an HDInsight cluster from PowerBI Desktop. I have no problem connecting to the database, and after I select the tables to load it gives me the option of loading the data or doing using DirectQuery. If I select DirectQuery, the connection complains that the data contains null values, which it does. This isn't a problem if I load the data into PowerBI, but I'd rather use the DirectQuery if possible.

[pqian · Accepted solution]

Hi @clackner

DirectQuery for Spark is still in Preview phase. Right now DirectQuery only officially support SQL server and SAP Hana.

Right now, please use Spark in Import mode.

[clackner]

Hi @pqian,

Thanks. Is there a way to give feedback on the preview mode?

[pqian]

Yep, preview features are regular things that we take seriously. It's more for you guys as a warning that we haven't actually completed the full cycle for the feature (which may mean more bugs, less performance, etc)

All feedback channels are open for preview features. In fact, your thread here made me aware of an issue that need to be addressed.