[jBPM Users] - Re: workflow design about wait states
by mmusaji
"kukeltje" wrote : So you want a front-end thread to be blocking for (mostly) parallel executed jobs in the backend? Ar you sure? How long will these jobs be running? Is the process fully ended after these jobs like in your definition that was posted?
|
| So basically http://www.eaipatterns.com/BroadcastAggregate.html is what you want?
|
The jobs will be running for different periods of time but anywhere between 2 and 8 seconds. But there could be problems with comms etc and it could take longer or we could retry etc so the job would take longer in them circumstances. And yes... unfortunately we are sure! lol.
But yes that pattern looks spot on. Is there an "Aggregator" of sorts in JBPM? Or would that be the Join node in a way.
"kukeltje" wrote :
| And I assume there are multiple return variables since having multiple parallel jobs working on the same variable is kind of dangerous.
|
Yes multiple return variables (each fork changing its "own" object) which is then read back at the front-end once its been updated.
View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4255509#4255509
Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4255509
16 years, 6 months
[jBPM Users] - Re: Exeception Handeling problem in JBPM
by kukeltje
You initial assumption is wrong:
anonymous wrote : In my JBPM approval process, I have a requirement that, if at any point of time JBPM exception happens during the process work flow., the control should delegate to a separate TaskNode called Admin node,
Because as you state yourself, this contradicts the documentation:
anonymous wrote : Note that the exception handling mechanism of jBPM is not completely similar to the java exception handling. In
| java, a caught exception can have an influence on the control flow. In the case of jBPM, control flow cannot be
| changed by the jBPM exception handling mechanism.
So think out of the box. A JbpmException should be handled by the delegationclass in which it occurs. What you can try (never did this, only did escalation with timers) is to raise an event, catch this event in your task and in the delegationclass in the eventhandler move the *token* to the corresponding node and move it back to the originating task when the admin task should be ended.
View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4255473#4255473
Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4255473
16 years, 6 months