![]() | |
![]() |
| | Thread Tools | Display Modes |
#1
| |||
| |||
|
#2
| |||
| |||
|
|
Hi, I have a set of KPIs (about 150 and it can grow). For each KPI i am creating a cube. At the end i am creating a virtual cube ouf of all these cubes. But this way the processing takes quite long. Is this a good/correct design? Or should i create less cubes (with more measures per cube), so the processing can perform faster. I have noticed from other designs that process less cubes with lots of measures processed faster. Thnx Stanley |
#3
| |||
| |||
|
|
Hi Stanley, Yes it would if the measures are coming from the same fact table then yes combine the cubes into one or a set of related measures. This way Analysis Services only has to query the data source once instead of once for every KPI cube. David Botzenhart "Stanley" <Stanley (AT) discussions (DOT) microsoft.com> wrote in message news:FB3830F3-6BE1-4365-B5D7-65BD148B9484 (AT) microsoft (DOT) com... Hi, I have a set of KPIs (about 150 and it can grow). For each KPI i am creating a cube. At the end i am creating a virtual cube ouf of all these cubes. But this way the processing takes quite long. Is this a good/correct design? Or should i create less cubes (with more measures per cube), so the processing can perform faster. I have noticed from other designs that process less cubes with lots of measures processed faster. Thnx Stanley |
![]() |
| Thread Tools | |
| Display Modes | |
| |