1.How will u implement drill-thru from powerplay to impromptu?2.How will u highlighting the data other than conditional formatting?3.What is 4-4-5 weeks in powerplay?Why r u using this?4.Why r u using multiple data source?and why cant u use single data source?5.How will u improve the performance of the cube?6.How will u relate the sources in transformer?7.Have u faced any errors while generating the models?8.What is uniqueness error?9.What is the use of Unique property?10.Why there is a property called move?what is its use?11.If u right-click one level u can see the properties?In that one drill thru property is there in source tab?What is the use of this?12.What is use of before rollup and after rollup?

jyothiph
Profile Answers by jyothiph Questions by jyothiph
Sep 2nd, 2006
6
2140

Showing Answers 1 - 6 of 6 Answers

dharani

Sep 8th, 2006

1.How will u implement drill-thru from powerplay to impromptu?

Ans:

In trsansformer right click on cube and click on properties

->in properties screen click on drill-thru tab add the impromptu imr report

Thanks

Dharani

dharanics

Sep 8th, 2006

4.Why are you using multiple data source?and why cant u use single data source?

Ans: when building the Cube it will take less time.
if u use single data source it will take more time

Thanks
Dahrani

dharanics

Sep 8th, 2006

5.How will u improve the performance of the cube?
Ans:
1.Delete the temp files
2.use flexing processing.
3.Use multiple data sources
Thanks
Dharani

dharanics

Sep 8th, 2006

12.What is use of before rollup and after rollup?

Ans
Before roll up:Calculations are calculated on source values when the cube is build
Afterrollup: Calculations are calculated on after building the cube.

Thanks
Dharani CS.

ripirr
Profile Answers by ripirr

Oct 10th, 2008

Why do we use multiple data sources?

Ans - It is a good practise to separate the FACT and Dimension IQDs, the dimension iqds execute quickly while the FACT iqd takes longer time to generate data/populate the cubes. For performance (cube build time) reasons it is good to have an iqd for each dimension table/view.

ripirr
Profile Answers by ripirr

Oct 10th, 2008

Try to limit the data in the iqds to the bare minimum, i.e., only fetch the data that is essentially required in your report via the iqds and check the 'auto-partition and passes' on the cubes. Higher partition size will build cube faster but will have slower access (by users). Have separate iqds to separate data fetches from multiple tables.