Global Recover/Termination Stage on Main Page

[jgregor4]

Is it best practice to have a Recover stage on the main page outside of any blocks that would link to an exception stage with no recovery or resume to handle unexpected exceptions. The result of the flow would be the process showing as terminated in the control room.

[John__Carter]

If an exception reaches the main page unhandled, then the process will terminate anyway. A common option is to do as you suggest and have a 'final' Recover to intercept a terminal exception, not in an attempt to solve the problem, but as a means of providing the chance to send a notification. In other words, the process is about to terminate, the exception is temporarily caught, an email notifying Admin that 'Process X on machine Y has terminated because...' is sent, the exception is re-released, and the process dies. Take a look at the template processes on this Portal for an example.

[Anonymous]

I was thinking about this same issue this morning. In addition to the above, a main page recovery stage is beneficial to perform ""Clean Up"" tasks such  as terminating any open applications before terminating the process.

[Denis__Dennehy]

I recommend all solutions should be based upon the Process Templates we provide.  Your retry loops within sub-pages should have 'clean up' logic to force items through the unexpected if possible, your 'Mark Item as Exception' sub page should have the logic to decide if the process should be terminated or to decide what to do if an application/system is down.