Forum Discussion
Speed Differences between Live Connection to Power BI Dataset and Imported Dataset
Hello amazing community!
I've been poking around for any info/tests related to my questions and haven't found anything yet. Would appreciate any insights!
In my head there should be (practically) no difference in speed between a report using import-mode for its dataset and a report using a live connection to a Power BI dataset. Is this correct thinking? There are many posts speaking about live connections to other Tabular services, but not comparing these two specifically.
One obvious implication is that I wouldn't want my "golden dataset" to feel really fast on one report and slower (DQ speeds) on others. Imported datasets are uploaded to the service (Azure DLG2). If the live dataset lives in the same place on the service (Azure DLG2) as the datasets which are imported, why would/should there be any noticible difference in speed?
If there is a speed difference, does anyone know how severe it is? Is a live connection to a PBI dataset just as slow (compared to import) as any other DirectQuery connection, or is it somewhere in the middle?
Many thanks!
1 Reply
Hi Anonymous
When using Import or LiveConnection to a Power BI Service dataset the speed should be fast either way.
The difference will be if you are using DirectQuery because that then has to go down the underlying data source. And if you are going via the Power BI Service dataset, there is that extra hop which would add a bit more time, which would make the report feel slower?
