As a coincidence, I'm also creating a dimensional data warehouse at the moment in Postgres and I'm also relatively new to Talend. I'm not sure if I understand the question correctly, but you can definitely as part of a job have Talend create tables in the tPostgresQLOutput component if they don't exist yet and indicate primary keys. I personally prefer to have the modeling tool create the DDL in advance and have Talend just populate the tables. In terms of populating the tables, especially when you're working with large dimensions (>150K records) and slowly changing dimensions, you want to make sure you use the ELT components for SCD and not the standard ones as they don't scale out well.
As a coincidence, I'm also creating a dimensional data warehouse at the moment in Postgres and I'm also relatively new to Talend. I'm not sure if I understand the question correctly, but you can definitely as part of a job have Talend create tables in the tPostgresQLOutput component if they don't exist yet and indicate primary keys. I personally prefer to have the modeling tool create the DDL in advance and have Talend just populate the tables. In terms of populating the tables, especially when you're working with large dimensions (>150K records) and slowly changing dimensions, you want to make sure you use the ELT components for SCD and not the standard ones as they don't scale out well.
