I have an issue with reporting requirements based on a cube where the end user would like to include attribute columns not currently included in the model.
As an example I have a customer dimension with the levels Manager, Sales Rep, Customer (customer being the lowest level). On a report the end user would like to see the Customer, contact, phone number and city. None of these last attributes of the customer are currently included in the model and they all exist on the customer master record. The report also has a whole serious of aggregated measures based on relative dates (YTD, Prior YTD, etc).
If I used a DMR, I could add the attribute fields, but I would lose the speed of the cube.
Any ideas on what to do here would be appreciated
Yes, it is possible though with limited no. of attributes.
Open the property of the level and you will see that you can add associated roles. Short Names, Long Names and Member description are 3 you could use easily and associate those with Contact, Phone No and City.
Thank you. That is what I have done up to this point, but I was hoping there might be a more elegant solution.
It is a bit of a myth that DMR would always be a lot slower than a cube. As long as the generated SQL is performing nicely the true bottleneck is the Cognos server needing to generate the Webpage. We found little performance difference between DMR and Powercubes with limited datasets. However the relative date dimensions are a different ballgame; this requires some imaginitive solutions within DMR