cancel
Showing results for 
Search instead for 
Did you mean: 

Backend application change monitoring

FélixGranados
Level 3
Hi all,

I would like to start a discussion about one thing that I guess most of us (if not everybody) suffers from time to time. Sometimes robot execution terminates because the backend application where the robots tries to execute the process (SAP, web application...) has changed and then we need to recalibrate the robot again.

The position we have in general about this issues is just reactive, we wait until the application change, the robot crash and then we fix it but I would like to have a more proactive approach. The options I see are the next:

1) Speak with IT and other stakeholders so we are informed about future changes: In few cases is an option but, in general, simply there are too many changes to use this option broadly 
2) To use a "monitoring robot" that test the applications before the the real robots are executed. Is an option but maybe too manual
3) To use a commercial tool design for this feature: This would be great but I´m not aware of any exiting tool

So what I would like to ask is, how do you face this situation? Do you use any of the previous options? any other option?

Best regards,
Félix

------------------------------
Félix Granados
------------------------------
10 REPLIES 10

The statement of "minor changes...hundred of times per day" jumps out to me as well, but it does make me wonder if it's more a result of the application being generated in a dynamic fashion and/or if the use of dynamic/wildcard attributes in the application modeler would get around this. I suppose it depends on how "minor" these application changes are.

Likewise, I would wonder if there are enough consistent identifiers available to toss out the filters causing issues. Sometimes an alternate spy mode (AA/UIA) can be more useful/reliable.

------------------------------
Ami Barrett
Automation Developer, Team Lead
Blue Prism
Plano, TX
------------------------------