I am a bit confused about the number of active instances of Entity beans. According to the manual the default locking policy is to allow only one active instance of an entity bean.
In my setup I have a BMP Entity Bean with a few set operations and one get operation. I use Standard BMP config with commit option set to B. The get method is marked as read-only and the set are marked for transactions.
1. First client looks up bean and gets data
2. Second client looks up bean and sets data
3. First client gets data
At this point the data is inconsistent. Browsing the log I can see that there is a second activation of the EB when the second client access the bean, even though it is in the pool.
The data inconsistency goes away if I merge the code for the two clients and executes it from one single client.
Anyone got a clue what I am doing wrong, or have any other generic ideas?
> Browsing the
> log I can see that there is a second activation of
> the EB when the second client access the bean,even
> though it is in the pool.
Doh, not what I meant. I can see that the there is
a second activation of the bean even though
the bean has not been passivated back to the pool.