Hiya,
I've implemented a mechanisms that gives you retries, async executions and also allows human interventions. 
You can take a look at the code here: 

https://github.com/Salaboy/human-task-poc-proposal/tree/master/executor-service

I will be working on improve what is there hopefully during the following month, but I believe that this project is exactly what you are looking for.

Cheers


On Tue, Nov 20, 2012 at 2:59 AM, <sbremal@hotmail.com> wrote:
Hello

I have a requirement for the following scenario: the jBPM workflow suppose to check the availability of various IT systems, it will also make changes to the systems. Each step is implemented as web service (implemented in Python).

It is expected that each step fails multiple times, and here comes the real issue. We need a smart user driven retry mechanism. The aim is that a failed step will generate a user task where it can be retried, skipped, aborted etc. (The workflow and further checks are suspended until a failed checkpoint step is remedied manually and is retried.)

How would you guys implement it?

I imagine a user task with an HTML form (error message, 3 buttons etc.). Where do I put the web service error handling? Where do I put the workflow signalling code? How do I make it generic so it 'guards' all web service calls in the entire workflow?

Thanks.


Cheers
B.

_______________________________________________
jbpm-users mailing list
jbpm-users@lists.jboss.org
https://lists.jboss.org/mailman/listinfo/jbpm-users




--
 - MyJourney @ http://salaboy.wordpress.com
 - Co-Founder @ http://www.jugargentina.org
 - Co-Founder @ http://www.jbug.com.ar
 
 - Salatino "Salaboy" Mauricio -