Page 1 of 1

handle different metadata using a single job?

Posted: Sun Dec 05, 2010 11:33 am
by vij
I have a requirement to create a job which can handle 3 different files, each one of them has different metadata and writes the data to Oracle tables?

Any help on this is highly appreciated.

Posted: Sun Dec 05, 2010 12:54 pm
by ray.wurlod
What's wrong with writing three separate jobs each with correct metadata?

Otherwise investigate an approach that uses Runtime Column Propagation.

Re: handle different metadata using a single job?

Posted: Sun Dec 05, 2010 11:44 pm
by sureshbabu
hi,

use schema file , and pass as dynamica parmetter

Posted: Mon Dec 06, 2010 2:23 am
by ray.wurlod
... and on the database side of the job?!!

Posted: Mon Dec 06, 2010 5:29 am
by Althaf6553
Try this appraoach
1> Import meta data of this sequential files into repository
2> Go to job and enable run time column propgation.Parameterize file name and table name.Enable run time column propogation in the general tab of properties window
3> Build a sequence job having 2 star loop activties and a job activity
In one start loop activity give sequential file name list and in the other the table name list (same as what is in repository) .Use list loop functinality of the start loop activity stage
4> run the job activity with in these 2 loops for file name pass Loop Activity1.Counter value and for the table name as Loop Activity2.Counter value .In the trigger give unconditional(Place the end loop appropriately)

This sequence will populate data from three different files into three different tables

Posted: Mon Dec 06, 2010 6:06 am
by samyamkrishna
use three seq file stage then use three modify stage or transformer stage to get them into same metadata,then funnel all the records and send to the database .....

Posted: Mon Dec 06, 2010 11:25 am
by naveen.p
Hi,

1. Create a three Schema Files for the three sequential files.
2. Parameterize the schema file and File Name Property.
3. Enable the Runtime Column Propogation at the Job Level.
4. Parameterize the Database Stage.
5. Enable the Multi-Instance Option at the Job Level (In case if the job is executed parallely).
6. Execute the Job using the required parameters.

Note: The above steps will be useful if there are no transformations to be applied on the data.

Any Suggestions is appreciated.

Naveen