Hi there,
meanwhile I succeeded in getting OpenJPA to work within my J2EE 1.4
container and
I was able to replace a CMP entity bean by a JPA Pojo and perform basic
operations
on it (insert/persist, query).
The hi-lo table sequence generator works fine, so I can avoid conflicts with
existing
generated PKs already in the database.
Before I can start the real world migration I just want to clarify an
architectural question,
your ideas would be welcome:
- I am using OpenJPA is "the persistence engine of choice" now
within our Java2EE 1.4 application
- all new persistent entities will be created using JPA
- all existing CMP EBs will be migrated/replaced by JPA in a longer
process (step by step as this is very complex)
- the application runs on 3 databases, one for production data, a
repository of additional stuff and a statistics module
My question is:
How would the most efficient way look like to integrate OpenJPA with three
databases within our application (which - from the
JPA point of view - means outside a container).
I want to avoid to invoke
emFactory = Persistence.createEntityManagerFactory("lbsims");
EntityManager em2 = emFactory.createEntityManager();
.
every time I need to perform a query of something else. Somehow there should
be a Util-Class that provides already
created Factories for all three databases (e.g. persistence units) within my
application. I already thought about implementing
this within an abstract base class of all my persistent pojos - but I think
this isn't a beautiful solution. At least because it
means mixing technical code with my domain object model.
How are people approaching similar problems?
Any ideas are appreciated!
HANS
===========================
virtually hanzz...
<http://hanzz.zapto.org> http://hanzz.zapto.org (personal)
<http://www.cse.dmu.ac.uk/~hansp> http://www.cse.dmu.ac.uk/~hansp
(research)
