Parallel Shared Container and Runtime Column Propagation

[wblack]

I have a generic question regarding parallel shared containers and runtime column propagation.

Do all input columns that go into a shared container's input automatically get passed through to the shared container's output? Does RCP have to be enabled to make this happen? RCP can be enabled on the shared container level or internally on each and every stage inside the container.

I know if you have N amount of columns that go into a shared container that the schema has to be on the output interface as well as any new columns that's introduced in the container. Does this sound correct?

[qt_ky]

RCP must be enabled in DataStage Administrator at the project level first, then you can enable RCP for new links in the job properties.

If you have an existing job without RCP and you want to enable it, then you will probably have to open one stage at a time to the stage's output column properties and check the RCP box. Maybe someone else knows some shortcuts.

Then all columns will automatically pass through a parallel shared container. On the output of the container, the RCP columns don't need to be listed.

[wblack]

I believe there's a shortcut for all stages inside the shared container. If you open the job properties and check or uncheck the option for RCP on new links it will prompt you at that time to apply the change to existing stages.

[qt_ky]

Thanks for the tip! I'll test it out next time I'm in Designer.