June 6, 2007 at 11:46 am
Is there any impact in defining fact as dimensions. I am planning to have a normalised structure for fact tables. For eg: Sales order specific information will be in sales fact tables and services specific information of a sales order is in sales services fact table.
There will be approximate 10 million rows of data in sales order fact and 30 -35 million rows in sales services fact.For analysing services sales_order_fact will become a dimension for services as well as sales_order_fact will act as an independent fact for sales orders.
My concern is if I define this normalised structure and defined sales order as a fact and dimension will there be any impact in terms of performance. I am using Sql 2005 DB and Analysis services for cube generation.
Please advice on this approach
June 13, 2007 at 7:37 am
Fact and dimension tables are totally different.
Fact table contains measures and different dimensions. Dimension tables just contains information about that particular dimension.
I don't understand your question about fact table as a dimension tables.
Can you give an example?
In Data warehouse, sometimes you cannot totally normalize. Maybe in this case you have to de-normalize for your dimension table.
Viewing 2 posts - 1 through 1 (of 1 total)
You must be logged in to reply to this topic. Login to reply