Hi vitalyla, I'm not sure if I followed your scenario exactly, but what version of OpenJPA are you using? As we were developing OpenJPA 2.x, we discovered a few inconsistencies with the clear processing, serializing, and detachment in general. These inconsistencies were documenting in this section of the user manual:
http://people.apache.org/~mikedd/nightly.builds/apache-openjpa-2.3.0-SNAPSHOT/docs/docbook/manual.html#migration_detach_behavior At the same time, I'm wondering about your getter method... I would not expect your getter method to be allocating storage for Child (new ArrayList...). I would expect that type of processing in your setter method. A few things to consider... Good luck. Kevin On Mon, Jan 14, 2013 at 7:27 AM, vitalyla <[email protected]>wrote: > Hi, > > I am facing an issue with detached entities when using > EntityManager.clear() > operation along with Local EJB. > Lets consider the case in which entity is returned from the query and later > detached by invoking clear() API: > > Parent > | > |_ Child (LAZY fecthed) > > implementation of getChild in Parent entity: > > Child getChild () { > > if (child == null) { child = new ArrayList....; } > > return child; > } > > The problem is that when Parent entity is detached through /clear()/ API it > is very common that at line when /child /object is returned it is still > NULL, regardless the fact that it was initialized a line before. > > But, when the same entity Parent is getting detached through > *serialization*, query exposed as remote EJB, then everything works as > expected. > > After checking an enhanced class it was found out that some state is > preserved in StateManager internal class when clear() is invoked but is > cleaned-up upon serialization. > > Please advise if it is a bug or this is how it works in: Local EJB + > EntityManager.clear() constellation. > > > > -- > View this message in context: > http://openjpa.208410.n2.nabble.com/EntityManager-clear-doesn-t-totally-detaches-an-entity-tp7582558.html > Sent from the OpenJPA Users mailing list archive at Nabble.com. >
