Aleksander, jBPM itself always relies on the DB for concurrency and transaction handling. If jBPM caches something, it means that information is considered static. Process definitions are static. When a new version of a process definition is deployed, it is stored NEXT to the old version. Both are kept in the DB. so the old version doesn't change. That is why jBPM can cache process definitions in memory. Whenever a process definition is being used, it is verified that it still exists in the DB. If not, it is removed from the cache. hth View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4255455#... Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&a...