Kernels¶

Here you are going to learn how to use default kernels and how to enable your customized environment based on e.g conda or virtualenv. Please contact support@dkrz.de for any issue or request.

System kernels¶

On Mistral, we povide the following Python kernels which are based on the Python module on Mistral:

Python 3 / Unstable
- it always point to the latest python module
- based on this new concept (see details)
Python 3 bleeding edge (using the module anaconda3/bleeding_edge)
Python 3 (using the module Python/3.5.2)
Python 2 bleeding_edge (using the module anaconda2/bleeding_edge)
Python 2 (using the module python/2.7.12)

Most of theese modules are based on conda, so you can use conda list to check all installed packages and their versions.

Even if Python is the language of choice, we also provide kernels for other programming languages:

R 4.0.3
Julia 1.5.2 (using the module julia/1.5.2-gcc-9.1.0)

Note

It is not possible for users to install or update packages. Sometimes you can use the --user flag to install new packages in your $HOME directory but there is no guarantee this will be available in the kernel.

Note

For testing new libraries/packages, we suggest that you use your own conda/env and you turn it into a kernel. If you/we think that a library/package is required by many users or interesting for data processing/visualization workflows, we will add it to one of the above kernels.

Note

Requests for new packages go through our system department and might take a time.

Use your own kernel¶

The following steps will show you how to create and enable your own environment.

Note

In order to use conda or virtualenv commands you need to load any available Python interpreter available. Recommanded is the module python3/unstable.

With conda

You need to create a conda environment e.g in your $HOME directory:

% mkdir $HOME/kernels
% conda create --prefix $HOME/kernels/your-kernel ipykernel python=3.x
% source activate $HOME/kernels/your-kernel
% python -m ipykernel install --user --name new-kernel --display-name="new kernel"
% conda deactivate

(Re)start the server and the new kernel should be available.

Kernel specifications are in ~/.local/share/jupyter/kernels/.

More details on kernels can be found here

With virtualenv

If virtualenv is not available, you have to install it first before trying the following steps. The best way to install virtualenv is with pip:

% python -m pip install --user virtualenv
% python -m virtualenv --system-site-packages /path/to/new-kernel
% source /path/to/new-kernel/bin/activate
% pip install ipykernel
% python -m ipykernel install --user --name new-kernel --display-name="new-kernel"

You can now add/install additional packages that you need in your new environment and then:

(new-kernel) % deactivate

Advanced¶

You can even go further with the configuration of your new kernel by updating the kernel.json. The content looks like this:

{
 "argv": [
     "/home/user/kernels/new-kernel/bin/python",
     "-m",
     "ipykernel_launcher",
     "-f",
     "{connection_file}"
 ],
 "display_name": "new-kernel",
 "language": "python"
}

It is possible to specify additional environment variables:

{
 "argv": [
     "/home/user/kernels/new-kernel/bin/python",
     "-m",
     "ipykernel_launcher",
     "-f",
     "{connection_file}"
 ],
 "display_name": "new-kernel",
 "language": "python",
 "env": {
     "variable": "value",
 }
}

Best practices¶

Where to install the new environment?

Depending on the number and size of the python packages installed in your new enviroment, disk usage in your $HOME can easily exceed the limit. Therefore, it is preferable to create conda/virtualenv environments in /work (in the corresponding project). In that case, creating a conda env for example in work:

% conda  create --prefix /work/project_id/$USER ...

Helper script for kernel.json

You can create a start-kernel.sh shell script and make it executable (chmod +x start-kernel.sh). Inside the script you can put all configuration you want in your new kernel. An example can be for example loading system modules. The sctructure of the script can be like this:

#!/bin/bash

source /etc/profile
module purge

module load netcdf_c/4.3.2-gcc48
module load python/3.5.2
python -m ipykernel_launcher -f "$1"

And the kernel.json:

{
 "argv": [
     "start-kernel.sh",
     "{connection_file}"
 ],
 "display_name": "new-kernel",
 "language": "python"
}

uninstall/remove a kernel
- jupyter kernelspec remove kernel
- delete the corresponding conda env if you don’t need it anymore

Spawner options Changelog