I am looking for a solution to reduce the size of Cognos cube. The cube in question is having daily granularity and has data for last 5 years and hence is very big in size.
I am looking for a quick solution where in, I can convert the cube from a 'Daily' to a 'Monthly' granularity. Is there something available out-of-the-box in transformer or somewhere else?
One of the issues is that some measures are calculated based on the number of days in the month. For Ex: number of employees in a month is a calculation based on the days of the month. Hence, the solution should be such that though the cube builds on data that is provided at 'day' granularity, once built, it only stores the 'month' granularity and discards/excludes 'day' granularity. Thus, saving space and enhancing performance.
Any suggestions would be welcome
And just to add, since the cube has two time dimensions, we cannot even look at time based partition as an option.
Do you have multiple "Alternate Drill Paths"? These are usually the biggest culprit for increasing cube size.
HI
Why is the size of the cube causing you issues? If it's because of build time or time required to open the cube in one of the reporting studios you can play around with that in the properties of the cube