Dataflow Merging & Privacy

[emarshall]

My understanding is that merging a query pointing at one data source (eg a SQL database) and another data source (eg a SharePoint folder) is risky (if options are changed to "Always ignore privacy level settings") as there is the potential for information to be leaked between sources. Chris Webb has documented this in his series below:

https://blog.crossjoin.co.uk/2017/05/24/data-privacy-settings-in-power-bipower-query-part-1-performa...

I am wondering whether there is still a potential for unwanted data transfer when dataflows are involved (eg I use a dataflow to connect to a SQL database, then I query the associated entity of that dataflow using Power BI Desktop, then I create another query in Desktop to files in SharePoint, then I merge the two queries in Power BI Desktop together).

Are dataflows an effective way of getting around this unwanted data transfer issue?

[StefanoGrimaldi · Accepted solution]

yes, if you create each entity in separate dataflow it should be fine, casue dataflows data isnt keep or manage on the source or the power bi service, its located on the azure datalake V2, and them you pulling information from that datalake via the dataflows, so if you have 2 dataflow indepent one from the other and quer the dataflows together on a new model it should be fine.