Best Practice: Combine data using Power BI or SQL statements?

[Anonymous]

Hi,

I'm relatively new to Power BI, dablling for about a year, seriously using for about 6 months. I've created many reports, but only 4 at this time that are useful and utilized. Now that I have these core reports built and published, my focus is on data architecture and efficiences, not only for these reports, but also future development.

My question stems around the best way to managae the data PowerBI consumes. The four reports that are in use, utilize a combination of MS Access and Postgres database connections. One of these reports comsumes over 700+k records of data. Currently all these records come from two Postgres views, each connected to PowerBI. Within the tool, I've used the "Append Queries as New" option to take the two queries and combine them as one. 

My question is, which is more efficient given that much data? Combining the data in Postgres and making one connection to PowerBI, or keeping it as is, two connections and appending to a new query within PowerBI?

Thanks!

[Greg_Deckler · Accepted solution]

In general, the more that you can push back to the server the better. This is a general rule. So, creating a View in Postgres that does is simply a UNION of your two other views is going to be the most efficient. You have those options in Power BI because most end users do not have access to the backend SQL database, but if you do, then that is generally the "best" way of doing it.