OWB - How to implement a type 2 slowly changing dimension with a hash function ?

> Data Integration Tool (ETL/ELT) > Oracle Warehouse Builder

1 - About

Slowly changing dimension is the ability to track change on a record from a data set.

How can we implement the type 2 to track the change and implement an history view.

The type 2 :

  • add a new record for update data
  • and flag the old record as the old data.

To achieve this goal, an hash function (as the function Oracle Database - ORA_HASH Function) is generally used. It permit to create a virtual primary key based on the value of each columns. If the content of a column change the hash value change and you are then able to target this changed record by using a SQL - Outer Join

Advertising

3 - Steps

3.1 - Add a hash column to the data set

With the function Oracle Database - ORA_HASH Function, add a formula column in the actual data set (the target) and the new data set (the source) which use all the trigger columns. (In this example, all)

Example :

ORA_HASH(INGRP1.DESCRIPTION ||
 INGRP1.TYPE_ID ||
 INGRP1.GROUP_ID ||
 INGRP1.HIERARCHY_ID ||
 INGRP1.EAN_CODE ||
 INGRP1.DIVISION_ID ||
 INGRP1.BASE_UOM)

3.2 - How to find the new/updated record ?

SOURCE.PK_HASH_VALUE = TARGET.PK_HASH (+) 
AND SOURCE.PK_HASH_VALUE <> TARGET.PK_HASH_VALUE

3.3 - How to find the deleted record ?

TARGET.PK_HASH = SOURCE.PK_HASH_VALUE (+) 
AND SOURCE.PK_HASH_VALUE <> TARGET.PK_HASH_VALUE