You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Sometimes workers will hang, despite having resource limits and other things. An outside script that kills things that have been alive for too long (wall-time) would be nice. It could run in a cronjob.
The text was updated successfully, but these errors were encountered:
This does not seem a problem anymore after dda0fed
A possible reason for "hanged" kernels was using the same ZMQ sockets from different threads leading to lost messages about killing kernels. We'll probably never know for sure ;-)
Sometimes workers will hang, despite having resource limits and other things. An outside script that kills things that have been alive for too long (wall-time) would be nice. It could run in a cronjob.
The text was updated successfully, but these errors were encountered: