Adding partitins to exsisting cube with no data in E table
We have a situation where we have to partition a cube which already has data, around 10 million records.
Now we are stuck with 2 questions, without answering them we are not able to proceed.
1 - The cube is already partitioned based on 0calmonth. From 1999 - 2004. But the cube is never compressed. Therefore no data is there in E table. So can we go ahead and add partitons to E table which is empty and later compress the data?
2 - Is 10 million records large enough for partitioning to show substantial improvement in performance?
Please assist me in answering these questions and build a better Data Warehouse.