Suppose you have a fact table in which you store orders from an e-commerce website.
There are various dates linked to a transaction:
* order date = the date the customer ordered something on the website
* due date = the date the package should be delivered at the customer
* ship date = the date the package left your warehouse
This means there are 3 links to the date dimension of the data warehouse.
An example of a fact line may be:
CustomerID OrderDate DueDate ShipDate Quantity Amount
14587 20140606 20140610 20140608 1 25.87
When you load this data into SSAS, you have two options.
1. You create 3 seperate dimensions (Order Date, Due Date and Ship Date), each based on the same date dimension of the data warehouse. This results in 3 cube dimensions (the dimensions linked to the measure groups in the cube).
2. You only create 1 dimension, simply called Date. You add the same dimension 3 times to the cube (so 3 cube dimensions). You rename them to Order Date, Ship Date and Due Date. In the dimension usage tab, you link the correct keys. For example, you link the Ship Date Key with the Date Key of the date dimension. The same for Due Date and Order Date.
The last option is the role playing dimension. You have only 1 physical dimension - the date dimension - but you link it 3 times to the measure group (the fact table). Depending on the link used, the dimension behaves differently, hence the name "role playing".
How to post forum questions.Need an answer? No, you need a question.What’s the deal with Excel & SSIS?
My blog at SQLKover
MCSE Business Intelligence - Microsoft Data Platform MVP