as for your point #2, the *only* time a call should ever get routed through writeLoad() with a null source is a collection cache. There is a hole in our test suite regarding optimistic caches in conjunction with collection caches because I had not intended to apply optimisitc locking to the collection cache regions. So this may be an issue; I'll add a test and see what happens... View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3957103#... Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&a...