We are building up the necessary tools to allow TAs to grant more time and have control over the resources of a course. Once this toolkit is available the documentation will go here.
What to Tell Your Students
The following information to students of a course (if relevant):
- For JupyterHub:
- The name of the prepared environment.
- For running jobs:
- The course tag
- The maximum job duration
- If jobs are preempted or not. If they are students must account for more time until they get the result.
- Paths to any prepared environment or installed software under
/cluster/courses and how to use it.
Students also need to know that the cluster is only for the course. If students runs out of GPU time for a course they need to contact a TA. TAs can ask our service desk
to increase time for given students.
Long running jobs can be canceled by more important jobs if the cluster is full. These jobs will automatically restart when there are enough nodes free again.
To avoid running the whole job from the beginning students need to know techniques to use checkpoints or save state in order to start from there when the job resumes.