DSXchange

Hi,

I'm a newbie about schema file.

As I can see in the Parallel Job Developer's Guide (5-28),

(note, however, that if you have defined
columns on the Columns tab, you should ensure these match the
schema file)

Does that mean the column name and other properties on the Columns tab must match those in schema file? But not all columns must exist in both scheme file and Columns tab?

Thanks!

All columns must exist in one or the other (or both). If runtime column propagation is disabled, all columns must exist in the link properties irrespective of what is in the schema file.

Just wanna clarify if my understanding is correct

e.g. Schema file has column "Car Code", "Car Name", "Car Description"


When RCP (runtime column propagation is disable

Column tabs of output link must have "Car Code", "Car Name", "Car Description", i.e. all columns.

When RCP (runtime column propagation is disable

Column tabs of output link can omit some columns. i.e. "Car Code", "Car Name" only is okay.

I do not use RCP, as it interferes with proper metadata management. Someone else will have to confirm this for you. It seems to accord with the description of RCP in the manual.

Even with RCP off you can work with a dataset and a partial schema definition. You can treat it like a relational table...

Once the dataset has been written, for reading it you can specify just the columns you need.

Klaus