I work for a large telecom company.
We are looking for a product to map Java and C++ objects into
a relation database (Oracle8) for the purpose of providing
object persistence. This product will be integrated into a new
product line that we are developing.
This Object-Relational binder/mapper middleware product should support the
following features:
- Support multiple concurrent top level transactions in a single process.
Our CORBA servers are multithreaded (thread-per-call).
The mapper must support multiple concurrent sets of
one transaction per database connection per thread.
- Be available for both C++ and Java platforms on Solaris/Sparc.
The Java version must be runnable with the Solaris Native Threads
version of the Java VM.
- Support transparent database read and write.
The in-core objects should be automatically read from the database
when accessed. It should not be necessary to call an explicit
function. At commit time, all modified objects should be
automatically written back to the database, without programmer
intervention.
- Support automatic concurrency control.
Object read and write locking should occur transparently as
objects are read and written in a transaction.
- Support normal Java JDK and user-defined classes as attributes.
The attributes of these mapped, persistent objects can be any
ordinary primitive type, array, or class. Having to use a
special String class, for example, is not acceptable.
Vendor supplied classes that add value are OK.
- Fully support normal object paradigms, such as polymorphism.
An excellent example of the kind of product we are looking for is
the "Ardent Java Bindings for Relation Databases" at
http://www.o2tech.fr/jrb/jrb.html. Unfortunately, this product appears
to be intended only for implementing single-user systems, as it fails to
meet critiera #1 above.
If you have a product or know of a product that meets the above
requirements, please post a response in this group or send email to
jrd_at_eskimo.com.
Thank you.
Jeff Demmon
Developer
Received on Fri Jun 19 1998 - 12:09:38 CDT