I have code that looks similar to the following:

Sesson session = ....; // get the session
User user = ...; // retrieve object from HttpSession
if (!session.contains(user))
{
session.lock(user, LockMode.NONE); // throws exception
}

How is it possible that I'm seeing a NonUniqueObjectException after checking via contains(...) that the instance is not already associated with the session?!?

was there an answer to this? I am having the exact same issue, and cannot figure out why this is happening. Any help would be great.

I've been able to recreate this by doing the following:

Foo contains a Set of Bar. (bidirectional)

* Get a Foo from the session.
* Evict Foo.
* Get a Bar that is associated with Foo from the session. This gets Foo as well.
* Check if Foo is known to the session with session.contains(). It's not.
* Try to lock Foo. NonUniqueObjectException is thrown, giving the name and ID of Foo in the error message.

When you load Bar with the associated Foo, a new persistent Foo instance is created in the session cache. Session#contains() does not match a detached instance of Foo -- like the one that you evict()ed -- since contains() uses JVM == rather than a comparison of persistent identity.

The purpose of contains() is to indicate whether a particular Java object is "persistent", i.e. whether or not that exact object exists in the session cache.

Unfortunately there is no interface for what you want to test: Is there a persistent identity equivalent object in the cache? (PersistentContext#checkUniqueness is not exposed in Session.)

I think there is a use case for exposing that test in the interface: If your session throws an exception you're supposed to discard it, so there ought to be a way that you can test if an update() or lock() is allowed without incurring the corruption of your session.

Is there is an answer to this question?
I.e. if there is a method to check if an hibernate session contains another object with class and identifier identical to that of a given one.

Thank you.

ttt

same problem ... did any of you get anywhere on this?

When I had a similar problem just now I tried the following workaround:

That got rid of the NonUniqueObjectException and I couldn't spot any unwanted side-effects.