I ran into an issue earlier when a test run was blocking a job, the process just completely stopped with no output. I ended up connecting through RDP and turns out I was missing a DLL (my fault entirely) and in the RDP session the job was blocked with an error popup.
Just wondering if anyone has run into similar issues and if there is a way to prevent these issues from blocking a job, or at least cause some output to appear in the log.

#1 Posted by Feodor Fitsner Support staff on 08 Apr, 2016 03:48 AM
I don't think it's that easy, but possible. Theoretically, before running tests you can asynchronously run some "monitor" checking for windows/processes and killing the test process if a window with specific title is found.
Another approach would be for AppVeyor to add some sort of "idle timeout" when the build is cancelled if there is no output for X minutes. Would that work?
Ilya Finkelshteyn closed this discussion on 25 Aug, 2018 02:05 AM.