Forum Discussion
How to Append Works ?
- Anonymous7 years ago
The loading from SQL should always be just be the two tables.
If you Append then the actual appending will be pushed to sql (though not a new table load)
If you append them as new, then Power Query will do the heavy lifting of appending them since query folding is broken.
But in the end, just two tables are loaded from SQL. The question becomes is it better to have SQL append them, or have power query. I'd say generally want to keep as much on the server as possible (as they are built for this) but depending on the size and cardinality, PQ might be just fine. If you are concerned I'd just run some test and see what shakes out. The good thing is that these tables are only updated at refresh time, which you can set for off-peak hours.
Thanks and thats a great info.. What I am actually trying to ask is...
I am concerned about the load going to my DW server due to Data Refresh in PBI Desktop. If i am having two independent query the two query will contact the DW server for data periodically. SO now the load is two for every refresh.
If I am appending these and creating a new one. Then when I refresh the data the load will be 3 or still 2?
I mean the appended new query will again contact my DW for data changes or will not?
Kindly help to understand.
The loading from SQL should always be just be the two tables.
If you Append then the actual appending will be pushed to sql (though not a new table load)
If you append them as new, then Power Query will do the heavy lifting of appending them since query folding is broken.
But in the end, just two tables are loaded from SQL. The question becomes is it better to have SQL append them, or have power query. I'd say generally want to keep as much on the server as possible (as they are built for this) but depending on the size and cardinality, PQ might be just fine. If you are concerned I'd just run some test and see what shakes out. The good thing is that these tables are only updated at refresh time, which you can set for off-peak hours.