Thanks Mauricio Salantino,
What my understanding is ...
a humantask defined inside the process, human task node is one of the node in the process flow, So every humantask belongs to one or more process flow. so the relatiship between human task and the process defined using the process id and task id. So when we store the process information in the database we can store the task information also in the same databse. So my final comment is all can handle in on epersintence unit using one schema. but the example and some jbpm documents shows diffirent two persistence unit and two schemas (like jbpm5 and task).
thanks,