I need to implement locking of a sort on my entity beans. The semantics of these locks are not straightforward; they operate at a "higher level" than either database record locks or the built-in ejb locks which JBoss uses to protect data integrity. They are intended to prevent users (who are manipulating a linguistics database) from stepping on each others' toes. In particular, these locks:
1. Span multiple transactions; they are explicitly requested and released by users. 2. Need to cover arbitrary portions of the data schema with a single lock request; i.e., getting a lock on one bean might entail that an entire graph of beans connected to this one should also be locked. 3. Need to be robust against partial failure. This seems like a good usage of the interceptor mechanism. That is, I would write a container interceptor following the existing design principles and insert it, I assume, late in the interceptor chain of the CMP 2.x entity bean container configuration. Then, all calls on the beans would have to pass through this interceptor before being executed and I could block those which aren't allowed (based on my arbitrary locking criteria). This seems far more efficient (and less error-prone) than checking for the necessary lock explicitly within every entity bean business method. Questions: 1. Is this a reasonable thing to do, or am I missing something? 2. Is there any more documentation of the interceptor architecture beyond the EJB container chapter of the admin book (and of course the source)? 3. Any particular hints for making this work right? Thanks for your time, Beau ------------------------------------------------------- This sf.net email is sponsored by: OSDN - Tired of that same old cell phone? Get a new here for FREE! https://www.inphonic.com/r.asp?r=sourceforge1&refcode1=vs3390 _______________________________________________ JBoss-user mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-user
