I am pretty new to jBPM, and tried to find solution in forums and wike without much success. Here is what I need: I have a single (reusable) process which processes a single file with a given name. To be able to run this process I define a superprocess with task node for user to provide the file name and this process as subprocess. This causes no problems. I also need to create another process which would take all file names from, say, a specified directory, create sub-process instances for every file, wait for all subprocess instances to finish, and report results from every subprocess to a user for further decision. I do not know how many files there are. This sounds like fork and join with variable number of tokens. How can I implement it? And how to pass parameters in this case? Also, because all these forked instances have the same process definition, I could try to run concurrent instances of a single transition (I found a reference that this is possible), but how to do this? And is there way to reflect this concurrency in GPD diagram? Thanks, Vlad View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4054101#... Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&a...