Error Handling in execution

Hello UiPath Masters,
It would be happen to know how to handle Errors. I found an option “ContinueOnErrpr=Ture” but it won’t much help for me.
I am looking some thing when any error occur while execution, UiPath should stop execution and automatically re start all.

Could you please guide me if any such options are available in UiPath.

Your inputs will be greatly appreciated

One way of doing it is by enclosing entire workflow inside try catch block .
In catch block set some variable(x) flag and use that variable outside the try catch block where you can validate the condition and loop the node back to starting point of try catch block .

PS:your workflow should be so robust and flexible that it’s should able to handle all the previous steps which left in the middle of the runtime(if application already opened ,
User already logged in, etc)


Thank you Dilip. Yes I agree with you. I suppose to have a check point for each step what I am performing. I do have check points for my script initially like detecting opened application and pick up the same application browser what ever the state it is.
In the main body of my script I suppose to keep check points so that execution will take care un for Sean errors can handle. This is old method which I used long back. I am looking for more advanced method if any such.
I have another thought to kick off slave UiPath script to monitor master UiPath script. It will always look for unexpected errors and if any such errors occour it will close error and re start master UiPath script. This Slave Script will loop depends on Master UiPath process status from windows task bar. If Mater UiPath process completed with out any errors then this slave also get stop. When we kick start Master Script, Slave also will start monitoring. This is just a thought but yet to see all fesabilitie… :)). This Master Slave concept only works if UiPath supports multiple instances are allowed to run in the same system at a time. If not then I will change my design.

It’s a very good input .
I assume you already tried with “retry scope” activity anyways that doesn’t follow what your said about " master /slave " concept .
@andrzej.kniola any thoughts on this ?

You can add retry scope by giving the condition on error so that it can perform the same operation in specified times.
This should be included for identified small flow of those failing certain times

Hey @rkelchuri

I have not read the other above stories but till so far what is good from my perspective best is to use “state machine” with try-catch inside states.

and trigger to next pr previous steps based on your exceptions. like if a state you have defined to open a SAP application and something went wrong or may selector exception will occur then your try catch inside that state sequence will throw an exception and you will use that exception to trigger it go back to again that step to again try to restart that SAP for number of times depends on your choice and conditions or if you do not wants to restart then just gracefully trigger it to final state and write log.



Only with High-Density robots, which seems like an overkill for this situation and also approaching it from wrong angle (IMHO).

If you’re handling workflow related errors correctly, and it seems you’re doing just that, things that might throw it off that come to mind immediately are:

  1. Network issues
  2. Robot service crash

If the monitoring is a robot, then anything not workflow related that will crash your main one will with high probability also crash the other one, or at the very least prevent it from running correctly.
If you want to set up monitoring on the same machine as the robot, I’d highly recommend it being a completely separate application/service, to minimize occurrences where everything crashes at once.

i just tried to run 2 UiPath scripts at a time to create master and slave concept but i it is not working.i am getting following error message.

Yup, like I said - it’s only possible for High-Density robots (which btw run on separate users, but share the machine).

I think you should rethink your strategy.

Let’s go back to the drawing board - what kind of errors do you wish to handle with the Master-Slave setup that you cannot handle within the workflow itself (or with the help of Orchestrator)?

Thanks for your inputs. Yes I saw multiple executions on the same machine with softoxxxx RPA tool which is compatible for master slave concept.

Basically handiling un predictable errors need error guessing. My intention of slave concept is more generic and dynamic. Which can be used for any work flow with out spending time on error handling while creating the work flow.

Here there is one more challenge on the design. While designing the workflow we need to have recover points. When work flow fails execution because of any un predictable errors, when we re start running RPA then Robot should be able to detect the application stage and should be able to bring back up to a recovery point and continue the task. Then only monitoring by slave concept will work.

How ever UiPath does’ t support to execute multiple copies on the same machine so we will hold discussion on this topic. If my R&D get success then I will share my thoughts further discussion.

Thanks to all for your valuable inputs. Greatly appreciated!

You could leverage 2017.1 new transaction progress field to set checkpoints and branch execution off of it. A top level Try-Catch would help with that as well, to make sure no exception crashes the robot and status + progress is always set.

Take a closer look at what Orchestrator offers currently. Only thing missing for a setup like this is auto-restart, but there are workarounds to that as well if you’re creative with your schedules.

hey u got anything from ur RnD. I also need something like that.
Hope if u could help…!!


Thin robots are available to execute multiple robots at a time in the same machine… but it’s in paid version. Contact in sales for thin robots. They will help you.

how to loop back to starting point? You mean CONTINUE activity?