The ultimate Fabric, Power BI, SQL, and AI community-led learning event. Save €200 with code FABCOMM.
Get registeredEnhance your career with this limited time 50% discount on Fabric and Power BI exams. Ends September 15. Request your voucher.
To complex query with multiple conditional joins; and several Sum or Count following group by after the join to extract more data; would you confidently arrange at Power Query or push back to SQL Server if any chance?
Kindly suggest with your best practices
Thanks and best regards.
Rich
Solved! Go to Solution.
Hi @rdnguyen ,
Absolutely 100% send to the SQL server instead of Power Query.
Almost all of the operations you've listed require Power Query to load the entire table into memory to perform, and therefore performance will be poor, especially if all these operations are done in the same query, one after another.
Nothing beats an SQL server at performing these types of operations.
However, with Power Query and the Power BI/DAX front-end you quite often don't need to do any of these things structurally at all. Properly constructed/optimised DAX on the VertiPaq engine will often outstrip or at least match SQL Server performance in many areas, depending on what you want to do.
Pete
Proud to be a Datanaut!
Hi @rdnguyen ,
Absolutely 100% send to the SQL server instead of Power Query.
Almost all of the operations you've listed require Power Query to load the entire table into memory to perform, and therefore performance will be poor, especially if all these operations are done in the same query, one after another.
Nothing beats an SQL server at performing these types of operations.
However, with Power Query and the Power BI/DAX front-end you quite often don't need to do any of these things structurally at all. Properly constructed/optimised DAX on the VertiPaq engine will often outstrip or at least match SQL Server performance in many areas, depending on what you want to do.
Pete
Proud to be a Datanaut!