implementation of type 2 scd

nkreddy_3 · Post by **nkreddy_3** » Tue Mar 07, 2006 2:03 am

hi,

i want to know how can we implement the type-2 SCD in datasatge server jobs.
you pls tell me the stages used while implenting the type 2 SCD

Sunshine2323 · Post by **Sunshine2323** » Tue Mar 07, 2006 2:22 am

Hi nkreddy,

Search on SCD will give you tons of information.

This link will get you started

ArndW · Post by **ArndW** » Tue Mar 07, 2006 2:23 am

Hello nkreddy_3,

A Type 2 SCD can be implemented in many ways in both the DWH as well as in the ETL tool.

Typically when you get a new column value in the SCD you will need to check it against it's value in the database; if it has changed you will need to update the old value and also to insert the new one.

Doing this lookup into the database can be slow, so often a hashed file is used to speed up this process. Thus you can successfully implement a simple SCD with just a transform and a hashed file stage (with one read and one update and one insert link going into it).

Depending on which tool and version you have, DataStage offers different stages for your use. For example there are CDC (changed data capture) stages available that can simplify your development and detection of changes in dimension values.

palaniappan · Post by **palaniappan** » Tue Mar 07, 2006 3:22 am

CRC32 function is the most used for CDC (Changed Data Capture). DS can be designed to support SCDs but it depends on how your Data Warehouse is architectured. And for your information, the next release of DS is said to have a separate stage that will handle SCD.

Thanks, Pal.