Spark HDinsight slow / 10 mill rows

[sse]

Hi. I am trying to load a quite big data set from Spark HDinsight (beta). In my experience its very slow (approx. 1 hour per mill rows). Any other experiences out there? I am using the Power BI desktop.

 

[ALeef · Accepted solution]

You could optimize the query if you only need to bring back certain columns or date ranges (rows generally).  Most of the time, I only need a few columns, or a certain date range, and that speeds up queries significantly.