Cube Problem

[abhishek.thakur]

I have a cube in which I am putting the data on a regualar basis by running a SSIS package.
I am currently storing my data in partitions of 15 days. I store the datetime upto which I have transferred the data into the cube in a seaprate table and after it I transfer the updated data from my fact table into the cube.E.g Suppose last datetime stored is 15 Aug ,I will transfer the data into the cube from the fact table which has been in the fact table after 15 aug.
But the problem I am facing now is that the newly updated data can contain some of the data of the previous data (of the same datetime)which has been moved to the cube earlier.

So now The problem is want to move only that updated data from the fact table which is not present the cube.
Please help .......any help is welcome....

[Island1]

This sounds very similar to a question from a few weeks back ...

Why are you doing this "manual incremental" approach to cube updates? If you do ETL each period to update the fact table, you simply reprocess the cube to perform a zero-based update. It's like you're trying to show only the "current period" in the cube or something ... If you have to pull certain date ranges from the fact table, is simply filtering on the "induction" side of Analysis Services out of the question?

If I can clearly understand the business need, I can perhaps make some suggestions.

Thanks!

Bill