One parametric job to load 100 tables from 100 files

[federico.l]

Hi,

I wish to create a Job that load a table from a file in a parametric way.

In other words, given the name of the file, the job should load the corresponding table in a straight way. (field names are identical between file and table and table has the same name of the file or anyway obtainable from the file).

Any idea?

Thanks.

[jerome_rajan]

1. Create a single RCP job with your source, target and any intermediate generic transformations.
2. Create a looped master sequencer with a delimited string of your file name/schema file name/ table name triplets in the start loop stage.
3. Use a user variable activity to extract each value into the 3 variables (i.e. File Name, Schema File Name, Table Name)
4. Pass these 3 variables as parameter values to your generic job in the sequencer.
5. The sequencer will run in a loop and load as many tables as you may have specified in the delimited string of triplets.

Hope this helps

[federico.l]

Thanks Jerome,

but is it not necessary to define metadata of destination table within the job?

[jerome_rajan]

No. The metadata for your flat files will be derived from the schema files that you will have to create and place in a location on your server.
The metadata for the target tables will be directly available to Information Server provided you supply the right credentials. This is how RCP works.

[WesL]

Here is a good tutorial on generic data processing.

[federico.l]

Many thanks for the very useful infos.