I am evaluating JBPM 5.4 with a PoC and am trying to realize the benefits of persistence that the engine does when a StatefulKnowledgeSession is created with an appropriate EMF. i.e a use case example of, say, a case when a workflow, while in its execution flow encounters a jvm shutdown. How is resuming from point of failure handled? Are there examples which clearly show this resume behavior - which API's to use etc..
Also, are there other jbpm project examples to get used to the tool and learn about the API. So far, I have access to the following:-
- https://github.com/droolsjbpm/jbpm/tree/5.4.x/jbpm-examples
- http://docs.jboss.org/jbpm/v5.4/userguide/
- Developer Guide
- http://salaboy.com/2011/02/04/jbpm5-community-training-%E2%80%93-module-2-the-bpm-practice/ and misc content there
And feel the need for more.
Did you check the jbpm5 developer guide? http://www.amazon.com/jBPM5-Developer-Guide-Mauricio-Salatino/dp/1849516448/ref=sr_1_1?ie=UTF8&qid=1377544316&sr=8-1&keywords=jbpm5 It is based on 5.4 and you can find the examples provided in the book here: https://github.com/Salaboy/jBPM5-Developer-Guide
Check this blog. I had used these as reference to creating samples for evaluation JBPM ( 5.1).
https://community.jboss.org/people/bpmn2user/blog/2011/12/18/jbpm5-web-example--local-human-task
Examples contains Human Interactions ( Human Tasks ) where persistence and fail over features of JBPM come into action.
NB: You may have to migrate these to 5.4
来源：https://stackoverflow.com/questions/18440523/jbpm-persistence-examples-resuming-a-workflow