can there be a problem in case two threads try to perform a jbpm workflow operation on the same token in case we use optimistic locking on all runtime data like tokens, process instance, taskinstance, variable instance... ? Jeff mentioned that this might result in a problem when one thread updates the token while the other thread only reads the token. In that case, hibernate's optimistic locking mechanism will not detect the problem. BUT... Isn't this the same as if the token-read-only-tx would have been performed before the token-update-tx in a serialized manner ? View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3972488#... Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&a...