Hi! How are you doing?
Currently, I am working on a project which aim at running pythonnotebooks on a supercomputer. The process is described below,
1. Logging in using an external webpage.
2. If login is successful, a script which allows to execute the pythonnotebook and working on it in some node will be sent.
The logging process is already working well. Now, I am looking for how can I run a job on the supercomputer using an API. Currently,
I am testing using Python and I would like to know if I am able to request resources using pyslurm.
The resources that I could request are the following:
#! / bin / bash
#SBATCH --partition = all
#SBATCH --time = 24: 00: 00
#SBATCH --nodes = 1
#SBATCH --job-name = "Jupyterhub_test"
#SBATCH --output = jupyterhub.out
#SBATCH --ntasks-per-node = 1
#SBATCH --gres = gpu: 3
cp /opt/jupyterhub/jupyterhub_config.py.
sudo jupyterhub
I do really appreaciate your help!