Sometimes runs hang due to internal signal states, or can appear to hang due to long running processes without terminal output.
A new endpoint could be introduced that creates a "core dump" of all sessions, state & signals etc to help diagnose hung scripts. This endpoint can be accessed before a run is killed and the output can be used alongside the console output to help investigate issues
The core dump could be produced automatically from a shutdown hook, too. That will let us investigate why qDup got stuck when the job is cancelled by a timeout.
Sometimes runs hang due to internal signal states, or can appear to hang due to long running processes without terminal output.
A new endpoint could be introduced that creates a "core dump" of all sessions, state & signals etc to help diagnose hung scripts. This endpoint can be accessed before a run is killed and the output can be used alongside the console output to help investigate issues