Skip to content
Go to file
Cannot retrieve contributors at this time
282 lines (185 sloc) 8.03 KB


Home_Icon Learning Center Home

Install Popular Data Science Tools on Atmosphere or Jetstream Instances


ez allows an "easy" installation of Anaconda based Jupyter notebook, Jupyter Hub, Singularity, and Docker.

This quickstart will cover specific commands for installing the following:

.. toctree::
        :maxdepth: 2

        EZ Installation of Popular Data Scientist Tools <self>
        Installation of RStudio on an Atmosphere Instance <rstudio.rst>
        Advanced Installation of Docker <docker.rst>


Downloads, access, and services

In order to complete this tutorial you will need access to the following services/software


We will use the following CyVerse platform(s):

Platform Interface Link Platform Documentation Quick Start
Atmosphere Command line (ssh) and/or Desktop (VNC, Guacamole WebDesktop) Atmosphere Atmosphere Manual Guide
Jetstream Command line (ssh) and/or Desktop (VNC, Guacamole WebDesktop) Jetstream Jetstream Manual XSEDE Portal

Atmosphere Images

``ez`` is deployed on all featured instances in Atmosphere and Jetstream

Get Started

Launch an Atmosphere or Jetstream instance.

Connect via ssh using a terminal.


Connect via the instance's web shell via the Atmosphere Instance's web page.

EZ Install Anaconda (Jupyter, Jupyter Lab, Jupyter Hub)

From your terminal session, you can install Anaconda Jupyter using the following
  • Jupyter notebooks with Python 3 (default)
  • Jupyter notebooks with Python 2
    ezj -2
  • Jupyter notebooks with R Kernel
    ezj -R or ezj -r
  • Jupyter Hub
Jupyter Notebooks

After the ezj installation, you will be provided a URL (e.g. http :// in your terminal session. Copy paste this into your new browser tab.


A Jupyter Notebook is running as an active process in the foreground on your Atmosphere instance. If you disconnect from your Atmosphere terminal session, the Jupyter Notebook will terminate.

One trick to keep the session running is to use a screen or tmux and disconnect the session before you close your browser tab.

To terminate your Jupyter Notebook, close the browser page with the Jupyter notebook interface. In your Atmosphere ssh session, press: control + C to terminate the Jupyter notebook.


Restart a new Jupyter session on a VM with EZ already installed

At the terminal command prompt retype ezj, this will restart the conda virtual environment and a new Jupyter notebook. Connect to the notebook using the URL as in the instructions above. Don't forget to use tmux!

Jupyter Lab (new)

Complete the ezj installation.

Jupyter Lab (Beta) is now included in the base Anaconda installation

To start a new Jupyter Lab session type:

jupyter lab

Install Jupyter Hub (new)

From terminal type ezjh

The installation may take up to 10 minutes to complete.

Once the install is complete, and a session is running, you'll get a URL address to the VM.

Copy paste the URL into your new browser tab.

You will be re-directed back to a CyVerse CAS service, log into your account.

The Jupyter Hub should now be loaded in the browser tab.

EZ Install Singularity

From your terminal, type the following command: ezs

You should see

Updating ez singularity and installing singularity (this may take a few minutes, coffee break!)
[sudo] password for YourCyVerseUserName:

Wait for the installation to complete.

Test Singularity

singularity run shub://vsoch/hello-world

EZ Install Docker

  1. `Connect to Atmosphere via ssh`_.

  2. From your connected session, you can install Docker using the following command:


  3. Test Docker

    sudo docker run hello-world


You need to use sudo permissions with Docker. After using the sudo command, Atmosphere will ask you for your CyVerse password for security purposes. You can remove the need to use sudo with Docker commands in the Advanced Docker Setup section.

Next Steps:

Some common next steps include:

  1. Installation of RStudio on an Atmosphere Instance
  2. Advanced Docker Setup
  3. Adding to Anaconda

Additional Information

Fix or improve this documentation

Home_Icon Learning Center Home