Can someone confirm the performance difference between a query that retrieves data from (1) a multi-provider or (2) an infocube?
The (1) example I'm thinking of is where actual data is stored in a cube, and planning data is stored in a second cube. In my example, the MP would join the actual cube and planning cube.
The infocube (2) would hold both actual and planning data, since in this example, the planning data was originally loaded into R/3, and thereby loaded into the info-cube with the actual data.
Which method will provide the faster query result ? A query written on the multi-provider, or on the basic cube ?
Regards,
Frederick