Hi Thomas. > The main reference is PropertyState, which is referenced by > PropertyImpl. PropertyState is kept in the transient space and / or > cache for a certain amout of time. I think it is not that easy to find > out when all hard, weak and soft references are removed, because this > depends on the virtual machine. That's why I have written the test > using the clearInUse method. To write a more realistic test case (that > works in the same way in all virtual machines) would be quite complex > I think. Do you think this is required?
I don't think that such a complicated test is required right now. The scenario I described in my last email happened while running a test that calls clearInUse() in runtime (and not while stepping thru the code). I think this issue is realistic, so I'm trying to find a way to notify the datastore about the removals from the weak hash map (so it can update the entry's timestamp). Ideally Session.save() could do this, so that there's no timeframe between the removal of those entries and the call to deleteUnused() in the GC. Regards, Esteban Franqueiro [EMAIL PROTECTED] Notice: This email message, together with any attachments, may contain information of BEA Systems, Inc., its subsidiaries and affiliated entities, that may be confidential, proprietary, copyrighted and/or legally privileged, and is intended solely for the use of the individual or entity named in this message. If you are not the intended recipient, and have received this message in error, please immediately return this by email and then delete it.
