I have been out of the actual development side of things for a few years, but was recently handed a large project that is going to be built with ADF and Toplink (12c). The project consists of a pre-existing database with almost 300 tables and views that will need to be mapped.
I decided to brush up on my EJB and JPA knowledge by running through the Oracle tutorial "Building a Web Application Using EJB, JPA, and JavaServer Faces". I decided to build the Java entities from some of my existing tables. This seemed to work well enough with the end result leaving me with a few single entity objects and several with composite key objects (w/ PK.java). Things start to go sideways when, as the tutorial directs, I try to build the Session Bean within the Entity Diagram. Basically, none of the entity objects created from the tables were visible in the wizard. So... I have been playing around with adding the entity objects to a persistance.xml and/or one or more orm.xmls so they would appear in the Session Bean wizard; however, all this has me doubting if I am using the correct approach.
Back 2007, I built a smaller, but similar application using ADF and Toplink (10g). Instead of "Java entities from tables", built "Java objects from tables", and I remember having no issues building the Session Bean; of course, there was no persistence.xml either... This smaller application is still live today, with regular upgrades, and is currently thriving in the AWS Cloud.
So I am looking for advice, and possibly guidance. For a project of this size, should I scrap the tutorial and build the new application the same way I did the smaller application all those years ago, or should I press forward with using the tutorial as a template to build the new project? Is there a better solution?
I would appreciate any insight anyone is willing to share.
Thank you.