SCD Stage Clarifications
Moderators: chulett, rschirm, roy
-
- Participant
- Posts: 80
- Joined: Sat Mar 18, 2006 10:27 pm
SCD Stage Clarifications
cc
Last edited by venkatvelpula on Tue Jan 24, 2012 8:50 pm, edited 1 time in total.
-
- Participant
- Posts: 54607
- Joined: Wed Oct 23, 2002 10:52 pm
- Location: Sydney, Australia
- Contact:
The model is, essentially:
The input has new data, including the business key. The business key is looked up against the Dimension table to return the corresponding surrogate key. This determines whether the dimension row yet exists and, if so, whether it is changed, so that the SCD part can be done. The Fact output includes the surrogate key value (whether generated or looked up) because it has to be a foreign key in the fact table.
Code: Select all
|
| Dimension
V
-------------> SCD Stage --------------->
| Fact
|
| Dimension
V
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
-
- Participant
- Posts: 80
- Joined: Sat Mar 18, 2006 10:27 pm
-
- Participant
- Posts: 54607
- Joined: Wed Oct 23, 2002 10:52 pm
- Location: Sydney, Australia
- Contact:
Easily fixed. Get yourself a premium membership. Then you will be able to read the entirety of Premium Posters' contributions. It's not expensive, less than 30c (Rs12) per day.
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
-
- Participant
- Posts: 3593
- Joined: Thu Jan 23, 2003 5:25 pm
- Location: Australia, Melbourne
- Contact:
I usually separate my dimension jobs from the fact jobs so when using the SCD stage to update a dimension I write the fact rows to a copy stage with no outputs to dispose of those rows. I then process just the dimension output rows. When most data marts have 10-12 dimensions it can be very difficult to create a single DataStage that updates all dimensions and the fact table in one pass - the job becomes very large and complex and harder to support. So I use a bunch of dimension jobs with SCD stages and a single fact job.
Certus Solutions
Blog: Tooling Around in the InfoSphere
Twitter: @vmcburney
LinkedIn:Vincent McBurney LinkedIn
Blog: Tooling Around in the InfoSphere
Twitter: @vmcburney
LinkedIn:Vincent McBurney LinkedIn