1 2 3 The result is ALL Sales Territory Country
Source Table GROUP BY Expression to evaluate
Compute sub-totals at the Year level
Optimizing means using the SE and avoiding FE
28
Catches events from SSAS Queries Events Query Processing
EVALUATE CALCULATETABLE( SUMMARIZE( 'Internet Sales', Geography[State Province Code], "Sales", SUM( [Sales Amount] ) ), FILTER( Customer, Customer[Last Name] = "Anand" ) )
Slower queryOptimized query Always put extended columns in ADDCOLUMNS parameters. Just use SUMMARIZE for its group by and when the evaluation context of SUMMARIZE is required (i.e. when a TOPN has to be executed for each group)
Using one SUMX, the optimizer pushes the calculation in VertiPaq
These two measures are slow to compute The filter is applied after computation
44 This query runs 7 times faster, even if it computes [Sales] twice
If you heavily hit Formula Engine, the cache will not be really useful.
Windows Azure