Is there a way to parse ALERTS in Appian and send that information to another sy

Here is an example on how you can handle this scenario: forum.appian.com/.../74931. As you mentioned before, once the node is paused by exception the flow stops and cannot continue. Therefore what you can do is:

1. Create a process variable to act as a success flag for those critical nodes or for the ones that fail the most, you will need one flag per critical node. This flag will be set to "true()" in a custom output in the critical node.
2. At the same level of the node, activate a timer that will wait, let's say 15 mns (assuming your node shouldn't take more than 15 mns to complete).
3. This timer should be connected to a gateway to review the value of the success flag
4. If the flag is true() then just go to an end event to complete the check
5. If it's not true it means the node didn't complete, therefore the flow needs to be redirected to an AND gateway connected, at the same time to the same node to restart it and to a Send E-mail node in case you want to notify somebody that the restart logic is going to be triggered.
6. For this to work gracefully the critical node (e.g. Call Web Service) will have the following options enable din the "Other" tab:

- Do not create if active instances exist.

As you can see this requires to add the failover logic to your model so this should be added just to the nodes that are more critical and that you see failing the most due to external factors.

1. Attached is an example application. In order to understand the explanation is important for you to LOOK AT THIS MODEL
2. The model is called "Exception Handling" under "Process Models" - "Exception Handling"
3. As you can see I have a script task that fails (because I forced a division by 0), at the same time I activate a timer to check the value of pv!success (which is set as a custom output of the script task called "Point of failure").
4. If pv!success=true() nothing to do, but go to an end event
5. If not, then go to start a new instance of the script task and notify about the error to somebody (optional)
6. The script task "ignore" is specific to this example and allows me to fix the division by 0 so the next time the script task works. Obviously you don't need this.

Here is an example on how you can handle this scenario: forum.appian.com/.../74931. As you mentioned before, once the node is paused by exception the flow stops and cannot continue. Therefore what you can do is:

1. Create a process variable to act as a success flag for those critical nodes or for the ones that fail the most, you will need one flag per critical node. This flag will be set to "true()" in a custom output in the critical node.
2. At the same level of the node, activate a timer that will wait, let's say 15 mns (assuming your node shouldn't take more than 15 mns to complete).
3. This timer should be connected to a gateway to review the value of the success flag
4. If the flag is true() then just go to an end event to complete the check
5. If it's not true it means the node didn't complete, therefore the flow needs to be redirected to an AND gateway connected, at the same time to the same node to restart it and to a Send E-mail node in case you want to notify somebody that the restart logic is going to be triggered.
6. For this to work gracefully the critical node (e.g. Call Web Service) will have the following options enable din the "Other" tab:

- Do not create if active instances exist.

As you can see this requires to add the failover logic to your model so this should be added just to the nodes that are more critical and that you see failing the most due to external factors.

1. Attached is an example application. In order to understand the explanation is important for you to LOOK AT THIS MODEL
2. The model is called "Exception Handling" under "Process Models" - "Exception Handling"
3. As you can see I have a script task that fails (because I forced a division by 0), at the same time I activate a timer to check the value of pv!success (which is set as a custom output of the script task called "Point of failure").
4. If pv!success=true() nothing to do, but go to an end event
5. If not, then go to start a new instance of the script task and notify about the error to somebody (optional)
6. The script task "ignore" is specific to this example and allows me to fix the division by 0 so the next time the script task works. Obviously you don't need this.