Note: JupyterHub with JupyterLab Install using CondaWritten on April 17, 2020 by Dr Donald Kinghorn
This is a quick note about setting up a JupyterHub server and JupyterLab using conda with Anaconda Python.
A few weeks ago I posted Note: How To Install JupyterHub on a Local Server In that post I used the system Python3 and a virtenv with pip to install JupyterHub. That is a perfectly fine way to do the install but I think I prefer what is presented in this new post using conda.
If you want to do this install yourself I recommend that you read through the post linked above since it contains more detail and discussion.
Install conda (globally)
Install extra packages
I needed to add curl on my Ubuntu 20.04 server test system, you may need other packages.
sudo apt-get install curl
Add apt repo for conda
curl -L https://repo.anaconda.com/pkgs/misc/gpgkeys/anaconda.asc | sudo apt-key add - sudo echo "deb [arch=amd64] https://repo.anaconda.com/pkgs/misc/debrepo/conda stable main" | sudo tee /etc/apt/sources.list.d/conda.list
This installs conda (miniconda Python environment) to /opt/conda
sudo apt-get update sudo apt-get install conda
Setup PATH and Environment for conda when users login
sudo ln -s /opt/conda/etc/profile.d/conda.sh /etc/profile.d/conda.sh
Install JupyterHub with conda
Sudo to a root shell for the install and setup
source the conda env for root
Using the "dot" command,
Create a conda "env" for JupyterHub with JupyterLab and install
We will want ipywidgets installed here too.
conda create --name jupyterhub jupyterhub jupyterlab ipywidgets
Create JupyterHub config file
Note: you will need to use full path to jupyterhub executable in the env.
mkdir -p /opt/conda/envs/jupyterhub/etc/jupyterhub cd /opt/conda/envs/jupyterhub/etc/jupyterhub /opt/conda/envs/jupyterhub/bin/jupyterhub --generate-config
Set default config to use JupyterLab
sed -i "s/#c.Spawner.default_url = ''/c.Spawner.default_url = '/lab'/" jupyterhub_config.py
Use systemd to start jupyterhub on boot
Create a systemd "Unit" file for starting JupyterHub,
sudo mkdir -p /opt/conda/envs/jupyterhub/etc/systemd
sudo cat << EOF > /opt/conda/envs/jupyterhub/etc/systemd/jupyterhub.service
[Unit] Description=JupyterHub After=syslog.target network.target [Service] User=root Environment="PATH=/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/opt/conda/envs/jupyterhub/bin" ExecStart=/opt/conda/envs/jupyterhub/bin/jupyterhub -f /opt/conda/envs/jupyterhub/etc/jupyterhub/jupyterhub_config.py [Install] WantedBy=multi-user.target EOF
Link to OS systemd directory
sudo ln -s /opt/conda/envs/jupyterhub/etc/systemd/jupyterhub.service /etc/systemd/system/jupyterhub.service
Start JupyterHub, and enable it as a service,
systemctl start jupyterhub.service systemctl enable jupyterhub.service
That's it! You now have JupyterHub installed and running. It will be running on the ip address for the system you installed onto using port 8000. Any user with a login account on the server will be able to access JupyterLab and use the machine.
There will be a default Python3 env available but will likely want to add some other "env's" for users. You can add env's for all users as in the examples below.
System-wide kernels are located in,
Individual users can create their own custom env's using conda in their home directory. Users env's will be located in
Add some extra (system-wide) kernels for JupyterLab
Anaconda3 (full package)
sudo /opt/conda/bin/conda create --yes --name anaconda3 anaconda ipykernel sudo /opt/conda/envs/anaconda3/bin/python -m ipykernel install --name 'anaconda3' --display-name "Anaconda3"
TensorFlow 2.1 GPU
If your server has NVIDIA CUDA capable GPU's and the NVIDIA driver installed then you can add GPU accelerated env's.
sudo /opt/conda/bin/conda create --yes --name tensorflow2.1-gpu tensorflow-gpu ipykernel sudo /opt/conda/envs/tensorflow2.1-gpu/bin/python -m ipykernel install --name 'tensorflow2.1-gpu' --display-name "ThensoFlow 2.1 GPU"
PyTorch 1.4 GPU
sudo /opt/conda/bin/conda create --yes --name pytorch1.4-gpu ipykernel pytorch torchvision cudatoolkit=10.1 -c pytorch sudo /opt/conda/envs/pytorch1.4-cpu/bin/python -m ipykernel install --name 'pytorch1.4' --display-name "PyTorch 1.4 CPU"
Happy computing! and please stay safe, best wishes --Don @dbkinghorn
Looking for a GPU Accelerated Workstation?
Puget Systems offers a range of poweful and reliable systems that are tailor-made for your unique workflow.
Why Choose Puget Systems?
Rather than getting a generic workstation, our systems are designed around your unique workflow and are optimized for the work you do every day.
By keeping inventory of our most popular parts, and maintaining a short supply line to parts we need, we are able to offer an industry leading ship time of 7-10 business days on nearly all our system orders.
We make sure our representatives are as accessible as possible, by phone and email. At Puget Systems, you can actually talk to a real person!
Click here for even more reasons!