Long-Term Job

In view of the fact that the farm Golias is a large-scale system, it is not possible to guarantee one hundred per cent reliability. On the one hand, it is necessary to provide system maintenance (e.g. rebooting worker nodes due to security kernel updates). On the other hand, an unplanned outage may occure, for example as a result of a hardware failure. Both may lead to an unexpected termination of jobs. Hence, it is crucial, especially for jobs running several days or several weeks, to save your results regularly.

Please check your long-term jobs from time to time. If jobs stop running, delete them and free up allocated processors for other users.

Přejít nahoru