2020-12-20 17:22:39 UTC, Zain:
>
> Hi,
> i am new to jupyterhub and i am trying to setup
> 1. add multiple python kernel(2.7.x,3.6.x,3.7.x) to jupyterhub
> 2. setup user specific kernel - means i have 3 kernels
> PY3, Pyspark and R. how can i assign this 3 PY3,
> Pyspark and R kernels to User1, User2 and User3
> respectively on Jupyterhub.
>
> my default python version is 3.6.8 installed from python
> sourcefile, and i am able to run jupyterhub with python3.
> Python path: /share/apps/platform/ai/py36/3_6_8/bin/python3
>
> But now i am trying to add Python 2.7.16
>
> Python path: /share/apps/platform/ai/py27/2_7_16/bin/python
>
> # jupyter-kernelspec list
> Available kernels:
> ir /root/.local/share/jupyter/kernels/ir
> pyspark /share/apps/platform/ai/py36/3_6_9/share/jupyter/kernels/pyspark
> python3 /share/apps/platform/ai/py36/3_6_9/share/jupyter/kernels/python3
>
> Please guide me to add multiple python kernels and
> also how to setup user specific kernels in Jupyterhub.
This is an answer to point 1.
One way would be to make a copy of your python3 kernel
$ cd /share/apps/platform/ai/py36/3_6_9/share/jupyter/kernels/
$ cp python3 python2
and to adapt the kernel.json by hand
$ vim python2/kernel.json
That is just a json file (a text file with special syntax)
which tells jupyter how to talk to Python and what
kernel name to display; adapt it so it talks to Python2
and displays Python 2 as the kernel name.
Hope this works for you. Kind regards, --Samuel