August 5, 2012 at 8:01 am
Hi,
I am kind of newbie on OLAP and I have some questions.
I have about 6 tables in my OLTP database that gets loaded overnight from different sources, to track the user activities in our distribution center; how many packages a user picked, packed or shipped or stocked in a particular day. So, here are the tables.
1) Picking Table: Column: UserID, ReportDate, PickCount
2) StockingTable: Column: UserID, ReportDate, StockingCount
3) ReceivingTable: Column: UserID, ReportDate, ReceivingCount.
And so on….
My question is; I want to develop a single cube that would contain all the counts (of picking, stocking, receiving…e.t.c) for all the users….The issue is there is no relations between these tables whatsoever…..They are independent of each other….But they will be related to TimDim table based on the ReportDate field.
And the measures (PickCount, StockingCount, ReceivingCount…) are stored in the same table…..In other words, both dimensions (ReportDate, user) and measures (PickCount, ReceivingCount..e.t.c) are in the same table….
Is a cube development possible in such scenario where no relation exists? Do we need to manually add the relations? Relation between the existing fields (userid, reportdate) or do we need to add new key into the relations?
What would be the best way of doing this?
I would appreciate if you can shed some light on this.
Thanks,
SQL_Surfer
August 6, 2012 at 6:20 am
This process is doable, although you may want to explore some design information on the web from the two many design patterns: Kimball is one and Inmon is the other.
Specifically, you can have multiple fact tables in a single cube and you can have "fact" dimensions, although I have found such a setup is not very performant. You may want to explore moving the data into some data warehouse tables which more closely match you desired end results.
Viewing 2 posts - 1 through 1 (of 1 total)
You must be logged in to reply to this topic. Login to reply