![]() | |
![]() |
| | Thread Tools | Display Modes |
#1
| |||
| |||
|
#2
| |||
| |||
|
|
I have a cube with a measure that indicates a maximum capacity. It can be summarized along all the dimensions, except for the time dimension, on which it should be aggregated by taking the maximum value. To get this done, I've used a custom rollup formula, something like MAX( time.currentlevel.children ). The performance of the cube after processing is very poor, in fact, I havent even been able to check if this produces the right results for me. I've tested performance with the 'browse data' function in the MAS Enterprise manager. The cube has 3 other dimensions that are rolled up by sum (plain rollup). Is there something wrong with my custom rollup formula, or is the bad performance what I should expect.? |
#3
| |||
| |||
|
|
I have a cube with a measure that indicates a maximum capacity. It can be summarized along all the dimensions, except for the time dimension, on which it should be aggregated by taking the maximum value. To get this done, I've used a custom rollup formula, something like MAX( time.currentlevel.children ). The performance of the cube after processing is very poor, in fact, I havent even been able to check if this produces the right results for me. I've tested performance with the 'browse data' function in the MAS Enterprise manager. The cube has 3 other dimensions that are rolled up by sum (plain rollup). Is there something wrong with my custom rollup formula, or is the bad performance what I should expect.? |
#4
| |||
| |||
|
|
"Tjibbe Chris Kuipers" <asdf@a.c> wrote in message news:3f847393$0$58698$e4fe514c (AT) news (DOT) xs4all.nl... I have a cube with a measure that indicates a maximum capacity. It can be summarized along all the dimensions, except for the time dimension, on which it should be aggregated by taking the maximum value. To get this done, I've used a custom rollup formula, something like MAX( time.currentlevel.children ). The performance of the cube after processing is very poor, in fact, I havent even been able to check if this produces the right results for me. I've tested performance with the 'browse data' function in the MAS Enterprise manager. The cube has 3 other dimensions that are rolled up by sum (plain rollup). Is there something wrong with my custom rollup formula, or is the bad performance what I should expect.? How many levels Time dimension has ? You might be better of by defining cell calculation with scope on all levels of Time dimension except the last one with the expression of MDX(Descendands(Time.CurrentMember, Time.LastLevelOfTimeLikeDays)) |
#5
| |||
| |||
|
|
- How/Where would I define cell calculation? |
![]() |
| Thread Tools | |
| Display Modes | |
| |