Hi,
I would like to ear from you about best practices to handle exceptions in workbench processes.
As far as I understand, one can handle exceptions only in some services. Also, each kind of exception must be handled, one by one (ex.: one route from the service that throws the exception by each exception).
This procedure as a lot of limitations: first, the services that expose/throws exceptions only throws some kind of exceptions. So, it is possible to make a route from all the exposed exceptions, but then get an unhandled error. Second, the process become too messy. Third, it seems we can´t garantee a fixed response message (in case of error), because we can´t control all the exceptions.
So, my question is: Am I right? Is it really true that we can´t catch all kind of process exception, for at least control the message we sent to the calling application?
If this is not possible, I really dont´understand how one can use this kind of processes in a production environment... So, I hope Im wrong :)
thank you
I would like to ear from you about best practices to handle exceptions in workbench processes.
As far as I understand, one can handle exceptions only in some services. Also, each kind of exception must be handled, one by one (ex.: one route from the service that throws the exception by each exception).
This procedure as a lot of limitations: first, the services that expose/throws exceptions only throws some kind of exceptions. So, it is possible to make a route from all the exposed exceptions, but then get an unhandled error. Second, the process become too messy. Third, it seems we can´t garantee a fixed response message (in case of error), because we can´t control all the exceptions.
So, my question is: Am I right? Is it really true that we can´t catch all kind of process exception, for at least control the message we sent to the calling application?
If this is not possible, I really dont´understand how one can use this kind of processes in a production environment... So, I hope Im wrong :)
thank you