Uploaded image for project: 'Red Hat Process Automation Manager'
  1. Red Hat Process Automation Manager
  2. RHPAM-4498

Boundary signal not triggered without providing process instance id

XMLWordPrintable

    • Icon: Bug Bug
    • Resolution: Cannot Reproduce
    • Icon: Major Major
    • None
    • 7.12.1.GA
    • jBPM Core
    • None
    • False
    • None
    • False
    • Hide

      Workarounds:

      1. Deploy application with RHPAM 7.11.1
      2. Include the process instance id in the signal API request.
      Show
      Workarounds: Deploy application with RHPAM 7.11.1 Include the process instance id in the signal API request.
    • Hide

      To reproduce the error the application team does the following:

      1. Build and deploy a RHPAM container pod with a process definitions that uses dynamic boundary signals (similar to RHPAM-3603). The signal name is typical 20-50 characters long.  

      2. Start a new process instance, for example:

      curl -X POST "http://localhost:8080/kie-server/services/rest/server/containers/test-container/processes/boundary-signal-test/instances" -H "accept: application/json" -H "content-type: application/json" -d "{ \"signalName\": \"mySignal_RA12345\"}"

      3. Progress process instance to human task with boundary signal. 

      4. Trigger a signal, for example:

      curl -X POST "http://localhost:8080/kie-server/services/rest/server/containers/boundary-signal-test/processes/instances/signal/mySignal_RA12345" -H "accept: application/json" -H "content-type: application/json" -d "{}"

      Observed behavior:
      Kie-server console doesn't show any logs regarding the execution of the signal API.
      The process remains at the human task.

      Show
      To reproduce the error the application team does the following: 1. Build and deploy a RHPAM container pod with a process definitions that uses dynamic boundary signals (similar to RHPAM-3603 ). The signal name is typical 20-50 characters long.   2. Start a new process instance, for example: curl -X POST "http://localhost:8080/kie-server/services/rest/server/containers/test-container/processes/boundary-signal-test/instances" -H "accept: application/json" -H "content-type: application/json" -d "{ \"signalName\": \"mySignal_RA12345\"}" 3. Progress process instance to human task with boundary signal.  4. Trigger a signal, for example: curl -X POST "http://localhost:8080/kie-server/services/rest/server/containers/boundary-signal-test/processes/instances/signal/mySignal_RA12345" -H "accept: application/json" -H "content-type: application/json" -d "{}" Observed behavior: Kie-server console doesn't show any logs regarding the execution of the signal API. The process remains at the human task.
    • ---
    • ---

      RHPAM 7.12.1 applications that extensively use BoundaryEventNodes signals triggered from an external system using the OOTB API that signals multiple process instances with a specified signal name are requiring processInstanceId.

      POST
      /server/containers/{containerId}/processes/instances/signal/{signalName}

      Every time the signal API is used on an application using version 7.12.1 the API does not execute the signals unless the processInstanceId is included on the request. This issue does not occur if the application is deployed with RHPAM 7.11.1. This is not the intended behavior for the API as processInstanceId is an optional variable. 

              rhn-support-mputz Martin Weiler (Inactive)
              rhn-gps-micperez Michael Perez
              Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

                Created:
                Updated:
                Resolved: