Setting up JupyterLab

Overview

Teaching: 0 min
Exercises: 0 min

Questions

• What is JupyterLab?

• How do I setup JupyterLab on COLA computers?

Objectives

• Explain what JupyterLab is and why we are using it

• Explain how to setup JupyterLab on COLA servers

• Explain what you can do with JupyterLab

Why do we need to setup another program?

We have been using the Unix command line to type and execute our Python commands. This is not a nice interface for working, especially as our codes get larger. We will use a browser-based interface called JupyterLab or Jupyter notebook. Jupyter notebooks are also ideal for scientific research because they make it easy to document our work, include equations in our documentation, show our analysis results, and link to other resources like journal articles. Let’s set it up and see what it can do!

Setup JupyterLab on COLA Computers

1. Login to a COLA computer (X refers to the computer number)

$ ssh -Y -l YOURUSERNAME colaX.gmu.edu

2. Load the anaconda module

$ module load anaconda/3

3. Setup a password for Jupyter. This will be different from your COLA or Mason passwords.

$ jupyter notebook --generate-config
$ jupyter notebook password

4. Start the Jupyter server with the following command (which asks for port 8878)

$ nohup jupyter lab --no-browser --ip=`hostname` --port=8878 &

nohup means “don’t hang up”, i.e., keep the process running even if your terminal session loses its connection (handy to prevent your work being lost, and allowing you to pick up later where you left off). The & at the end puts the process in the background, freeing up your command line to be used.

Pay attention here, if port 8878 is busy you might automatically get assigned a different port

For example:

[I 19:01:58.026 LabApp] The Jupyter Notebook is running at:
[I 19:01:58.026 LabApp] http://cola1.gmu.edu:8879/

Check that your process is running (a good thing to do if you are not sure whether you already have a Jupyter lab session running on the server:

$ ps -u YORUSERNAME

You should see one and only one enstantiation of jupyter-lab running. If you see more than one, then you should kill all the duplicate processes and relaunch the Jupyter server. Specific processes can be killed by their process ID (PID), e.g.:

$ kill -9 11404

Kill your process with PID = 11404 (you can provide a list of process ID numbers to kill all at the same time). The -9 option means to kill with extreme prejudice!

5. Start the Jupyter client on your laptop

In a separate XQuartz or MobusXterm window, log into the same cola machine (colaX) again with the following command

$ ssh -N -L 8878:colaX.gmu.edu:8878 YOURUSERNAME@colaX.gmu.edu

Nothing will appear to happen after you type your password. This is OK. You now have the Jupyter client running on your laptop.

Again pay attention!

Make sure that in steps 4 and 5 same cola machine and port is being used.

6. Connect your Jupyter client to the Jupyter server on the COLA system

Open a web browser and go to http://localhost:8878 (again, make sure the port number is correct). You will be asked to enter the password you created in step 1. Jupyter will appear in your local browser.

What can go wrong here?

There are various errors that can happen that will not allow Jupyter to launch properly:

Traceback (most recent call last): File “/homes/sknapp4/.conda/envs/aoes/lib/python3.6/site-packages/traitlets/traitlets.py”, line 528, in get value = obj._trait_values[self.name] KeyError: ‘allow_remote_access’

Solution: Try launching Jupyter with the following command

$ jupyter lab –no-browser –ip=’0.0.0.0’ –port=8878

If you get an Invalid Credentials error, check the messages on the screen where you launched jupyter lab Sometimes port 8878 is in use and it will find a different port. The error messages will tell you which port you should use.

These errors messages look like:

jupyter lab –no-browser –ip=`hostname` –port=8878 [I 14:12:20.887 LabApp] The port 8878 is already in use, trying another port.

Solution: Repeat step 6 using the reccomended port.

Your Jupyter client should continue running on the COLA server until you either kill the process yourself, or that machine gets rebooted. However, your client may stop if you lose your Internet connection, reboot your laptop, shut down your laptop, even close your laptop for more than a few minutes,… But all you have to do is relaunch your client (Step 5 above) and your Jupyter browser session should re-establish itself right where you left off.

8. In the upper right corner, you will see your current environment.

You may have others but we will start by using the default “Python 3” environment.

What do I do now that JupyterLab is setup?

Now that you have completed the setup process, you need to execute steps 5, and 6 when you want to run JupyterLab. If the cola server was rebooted, you will need to execute steps 4, 5, and 6.

Test that you can login and launch JupyterLab

Logout of COLA computers. Login again and launch JupyterLab

Key Points

• JupyterLab is an interactive Python interface that is ideal for scientific work

• Launching JupyterLab is a multi-step process

• Launch the server on a COLA computer jupyter lab --no-browser --ip=hostname --port=8878

• Login to COLA computer with special login line: ssh -N -L 8878:colaX.gmu.edu:8878 YOURUSERNAME@colaX.gmu.edu

• Open JupyterLab in your web browser: http://localhost:8878