• It is an active transformation to sort the data in ascending or descending order.
• It has to be used before Active Transformations to improve the performance.
• It can be used to remove the duplicate rows.
Ports: Input, Output, Key, Direction (ASC or DESC)
When the Key port is selected, respective direction port is enabled.
• Sorter cache (Separate memory)
• Work Directory (Temp directory)
• Tracing Level (Common for all the transformations – it is normal by default)
• Treat NULL rows low (unchecked) (Bring the NULL rows down, by default they are up in the result set)
Define Source (EMP)
Define Target (DIM_EMP)
Create Mapping (M_SORT)
Drag the Source (EMP) and Target to the Mapping Designer Workspace.
From the Transformation Menu select Create Sorter
Give the name T_SORT and click on Create and Done.
A blank transformation object appears on the screen.
Drag the required ports to the Sorter.
Edit the Sorter and in the Ports tab do the following settings:
Check the Sal and Deptno as Key and choose the order (direction – asc or desc).
Our objective is to display the department wise salaries in descending order.So change the order of the columns (make dept appear first then salary)
Click on Apply and OK.
Project the required ports from SORTER to the Target.
Go to Repository menu and Save.
Now Create the Session (S_SORTER)
Create the workflow (WF_SORTER)
Start the workflow.
Preview the data.
The target table is loaded with the sorter data.
Suppose we just want to know the salaries of deptno 10 then we need to filter out the records.
Note: To enhance the performance, first apply sorter and then filter transformation .
In this case the time elapsed is 3 seconds.