Posted on June 17, 2011, in BI Theory and Best Practices, Notifications and Updates, OBI EE Development, Repository and tagged , , , , , . Bookmark the permalink. 11 Comments.

  1. Great presentation. I appreciate the Canonical Time dimension modeling 😉


    • Thanks Nico. BTW, I’ve been heavily using your post on 11G installation as my guide for various installations – well done.
      Jeff M.

  2. Excellent Post Jeff. I’m sure this gets saved in many many desktops as a reference guide.

  3. Saichand Varanasi

    Nice article….

  4. I attended Jeff’s presentation in ATL and give it a definite AAA rating. Well done, Jeff.

    Also liked your suggestion for

  5. I like both “Combo Tables” and “Canonical Time” topics. Very exciting. However I ran into an issue on how to combine the two techniques together.

    The case is:
    1. sales fact table is mixed table (e.g. including non-numeric attributes order_type_code, transaction_type,etc) so we want to move the non-numeric attributes to Dim table ;
    2. sales fact table has multiple date columns (e.g. ship_date, cancel_date, order_dat,etc) so we want to take use of carnonical time model.

    We created multiple aliases for the sales fact for carnonical time model. The problem is how to move the “order_type_code, transction_type” to a dimension table to make it conformed to all sales fact aliases table? Is there a solution?

    Thanks in advance.

    • Sounds like you need to map each of the aliases to the logical dimension for your order_type_code. That should do the trick. The fact will drive which of the dimension LTSs to use, as each one only works with one of the Fact LTSs.

      Jeff M.

  6. Thanks Jeff for the reponse. I am not sure if i understand it correctly. According to the presentation, in order to handle combo tables, we need to use the same physical table for both Dim and Fact. Take alias Fact_Sales_Shipped for example, I need to drag it twice to BMM layer and join them, one uses as fact table and another one uses as dim table (for shipped_order_type_code). Then do the same thing for alias Fact_Sales_Canceled and I got the 2nd dim table for cancled_order_type_code. The thing is we don’t like multiple dim table for the order_type_code. Isn’t it good that we can have one conformed dim table for order_type_code for both fact_sales_shipped table and fact_sales_canceled table? Am I missing anything?

    To clarify, we don’t have dimension for order_type_code in data warehouse, it is in sales fact table (and that is why we are trying the “combo tables” in your presentation).

    thanks again!

    • well yes, a conformed dimension table is your best route. But if you can’t do that, you can have multiple aliases in the logical dim, each as a LTS from which OBI will pick only one at a time.

      • Jeff I think I got it and that is what I have done. Thanks for your help and I really like your smart solutions in the presentation.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: