3 Replies Latest reply on Feb 18, 2016 6:08 AM by jimmy001

    How to deploy my extended jBPM engine into jBOSS

    fardad.moosaei

      Dear All,

       

      I have a question regarding jBPM deployment into jBOSS. Actually I am doing my Phd and I have extended jBPM engine as a prototype to support the change management.

      To do so, I have connected the jBPM to the Neo4j graph data base and also have provided new swing interfaces to allow several actors run their tasks and signal change on their input data.

      But I need to validate my prototype by running it on several machines and let several actors execute their tasks in their own machine.

      -- I want to know how I can deploy my prototype into jBOSS with minimum effort (I mean manipulation)?

      -- do I need to develop new web interfaces?

       

      Thanks in advance for your help.

        • 1. Re: How to deploy my extended jBPM engine into jBOSS
          jimmy001

          Hi,

           

          a bit difficult to anwer without knowing more details. Perhaps I would go the route to let your swing client use the rest interface of the workbench.

          Have you integrated the Neo4j database by developing your own "PlaceholderStrategyResolver" or by using service tasks?

           

          J.

          • 2. Re: How to deploy my extended jBPM engine into jBOSS
            fardad.moosaei

            Hi,

            Thanks for your reply. In fact, here is a small description of my work. you can also find in  attach the image of my  prototype architecture.

            In fact, I have a central graph data base which stores the process instance elements (human task, data, actor) as nodes and edges ( actor performs task, task has input and out put data).

            What I want to do is to have a master jBPM who distributes the work items through other actors (remote engines) and creates the graph representation of the process instance into graph data base. Just in the new interface, I allow the users to signal change on their input data. Then As a master engine, will receive the change request and traverse the graph data base to find the affected elements.

            the integration of Neo4j with jBPM has been done in the work item handler. I just support human task work item handler. for instance, when ever a user wants to finish his task, after calling the

            complete funtion of the work item manager, I go through the data base and find the corresponding task node and update the state of it as "completed"

            so my question is:

             

            -  How can I deploy my extended jBPM into jBOSS (with new workItem handler)?

            cmc.png

            • 3. Re: How to deploy my extended jBPM engine into jBOSS
              jimmy001

              Hi,

               

              "What I want to do is to have a master jBPM who distributes the work items through other actors (remote engines) and creates the graph representation of the process instance into graph data base. Just in the new interface, I allow the users to signal change on their input data."

               

              I must admit I do not understand what you mean (might be my fault). If you have a "master" jBPM engine does this mean you also have "slaves" (the remote engines)?