When a running robot crashes (in Design Studio or on a RoboServer), Kapow creates an additional dump file that is very helpful in investigating why the crash happened.
Note: If the error is seen in the Management Console logs, the folders below need to be located on the machine where the RoboServer runs (the RoboServer on which the robot was running when it crashed). When the error is seen in the Robot Messages log from the Management Console , right click on it and select "Goto Run" to see more information about this robot run; this will also contain the RoboServer on which the robot was running at the time.
When an "Internal error during robot execution" is encountered, the dump files are saved here:
- Windows machine:
where <username> is the user running the Kapow process (e.g. the user that runs the RoboServer service. Or the user running Design Studio)
- Linux machine:
Note: Starting with Kapow version 9.7, the folder is no longer named "KapowKatalyst", just "Kapow".
This folder might contain multiple folders; the one containing the dump file is the one created at the same time that you've seen the error (if there are multiple folders created at about the same time, check each of them). Copy the .dmp file that you find and attach it to your case on the Kofax Customer Portal.
In situations where dump files are not creating, you can use Microsoft's ProcDump utility to hook into the process and generate one when the process terminates.
- From an Administrative CMD prompt, extract ProcDump into a folder.
- Make sure the process that is crashing is currently up and running (e.g. KapowBrowser.exe when in Design Studio testing - the process name would be KapowBrowser.exe).
- Type: procdump -e -t -ma processnamehere.exe C:\Temp
- In the above example, C:\Temp is the destination path for the dump file and it can also have ProcDump utility running from there as well.
- Reproduce the issue and zip up the trace and submit as requested.
Keywords: Kapow robot crash, Kapow, Robotic Process Automation, DMP