Jupyter

From FarmShare

Revision as of 12:54, 5 January 2017 by Phn (Talk | contribs)
Jump to: navigation, search

Contents

Introduction

Project Jupyter evolved out of the IPython project (specifically the IPython notebook) with the goal to provide an interactive, web-browser driven, language-independent programming environment. Jupyter notebooks can be deployed on the FarmShare servers to enable an accessible, powerful, and persistent computational platform.

Features

At the end of this guide, the resulting Jupyter notebook will support:

  • Python, MATLAB, R, SAS, and Julia programming languages
  • an encrypted, token protected, and web-browser enabled programming environment
  • indefinite persistence of the Jupyter notebook server environment with simple weekly renewals (the maximum duration of Stanford Kerberos tickets)
  • file/data storage on the Stanford AFS servers (5GB user quota, Stanford-wide, automatic backups)
  • shared file/data storage to Class Disk AFS Space
  • easy deployment on any of the Stanford FarmShare systems

Overview

The guide consists of three sections: installation, Jupyter notebook server management, and client connection. In typical usage (i.e., once installation is complete and the Jupyter notebook server is running), only the client connection steps are necessary at every use.

Installation

Jupyter setup is best performed via the Linux console. This guide will be mostly step-by-step, but general familiarity with Linux is helpful. Commands to be typed in are in bold. Useful tutorials are available as a free book, online tutorial, or video series.

SSH into FarmShare

ssh jane@corn.stanford.edu

substituting sunetid for jane.

SSH provides an encrypted remote shell into another system and is the primary way that Jupyter will be installed and accessed.

SSH into one of the farmshare computers. The corn cluster (i.e., corn.stanford.edu) is a good choice. Windows does not come with an SSH client, but putty is a a free and lightweight SSH client for Windows. Mac and Linux have ssh clients accessible via the terminal.

Bind to Jupyter virtual environment

corn99:~> bash
jane@corn99:~$ source /afs/ir.stanford.edu/group/bil/env/jupyter/bin/activate
(jupyter)jane@corn99:~$

These two commands switch the shell to bash and then update the environment to use a pre-built Jupyter installation. The prompt should have the (jupyter) prefix if the environment change is successful.

The pip packages installed for the Jupyter environment are stored in a requirements.txt file at /afs/ir.stanford.edu/group/bil/env/jupyter/requirements.txt. Additional packages can be installed upon request.

Setup encryption and default configuration

(jupyter)jane@corn99:~$ /afs/ir.stanford.edu/group/bil/env/scripts/jupyter_config_wrapper

This command calls a script that will create encryption keys for the Jupyter notebook and configure the jupyter config file to use these newly created keys. Note, this will modify existing Jupyter notebook config files (if they exist), but will not overwrite any parameters that have changed from their default values.

Installation is complete after this step is successful. Log out of FarmShare.

Jupyter notebook server

Jupyter notebook will be run within a persistent virtual terminal so it can stay running even after the user has logged out of FarmShare.

Create virtual terminal

This only needs to be performed once per FarmShare system (and after every server reboot).

SSH into a FarmShare system (e.g. corn). Make a note of exactly which corn server entered (e.g., corn14, corn22, etc).

corn99:~$ pagsh
sh-4.3$ kinit; aklog
Password for jane@stanford.edu:
sh-4.3$ tmux
(new blank terminal appears)
corn99:~$ bash
jane@corn99:~$ /afs/ir.stanford.edu/group/bil/env/scripts/keep_kerberos_afs
Run: export KRB5CCNAME=FILE:/tmp/.krb5_jane.tgt
jane@corn99:~$ export KRB5CCNAME=FILE:/tmp/.krb5_jane.tgt
jane@corn99:~$ source /afs/ir.stanford.edu/group/bil/env/jupyter/bin/activate
(jupyter)jane@corn99:~$ /afs/ir.stanford.edu/group/bil/env/scripts/jupyter_start

Substitute sunetid for jane in the export line (as instructed by the keep_kerberos_afs command). If done correctly, the output of this command will be

IMPORTANT: This Jupyter notebook is listening on TCP port 9876

[I 17:23:59.661 NotebookApp] Loading IPython parallel extension
[I 17:23:59.668 NotebookApp] Serving notebooks from local directory: /afs/ir.stanford.edu/users/j/a/jane
[I 17:23:59.668 NotebookApp] 0 active kernels
[I 17:23:59.668 NotebookApp] The Jupyter Notebook is running at: https://localhost:9876/?token=ba682763f27d8e2d59862badef28b0eaecb552529933176e
[I 17:23:59.668 NotebookApp] Use Control-C to stop this server and shut down all kernels (twice to skip confirmation).
[C 17:23:59.671 NotebookApp]

    Copy/paste this URL into your browser when you connect for the first time,
    to login with a token:
        https://localhost:9876/?token=ba682763f27d8e2d59862badef28b0eaecb552529933176e

The notebook will not work until a client connection is established. Copy the https://localhos... URL and save it for subsequent use. Also note the TCP port that the Jupyter notebook is listening on. Use Ctrl+b then d to detach from tmux. Logout of FarmShare.

Ctrl+b then d
(detaches tmux)
sh-4.3$ exit
corn99:~> logout

Note: the jupyter_start script accepts an optional port argument, specifying which port to connect to (e.g., jupyter_start 9876). This is useful in case a randomized port is not desired.

Renewing virtual terminal

This only needs to be done a week after creating the virtual terminal.

The virtual terminal created by the above instructions will persist on the FarmShare server for one week, after which its Kerberos tickets will expire and the Jupyter notebook will no longer function. To restore the notebook after this time, SSH into the same FarmShare system used to create the virtual terminal.

corn99:~$ tmux attach -d
Ctrl+b then c (to create a new window in tmux)
jane@corn99:~$ kinit; aklog
Password for jane@stanford.edu:
Ctrl+b then 0 (to return to the original window)

This will extend the terminal's credentials for another week. The same notebook URL can be used without interruption. Note, for subsequent renewals, use Ctrl+b then 1 to switch to window 1 instead of creating a new window. It is not necessary to create a new tmux window every time.

Client connection

Create SSH tunnel

Once the Jupyter notebook server is running, it is ready to accept client connections. For security, it only accepts connections from localhost (i.e., connections originating from the FarmShare system itself). Local connections can be created remotely through SSH tunnels. SSH into the same FarmShare system that the Jupyter notebook server virtual terminal was created on (e.g., corn14, corn22, etc) by using the server's FQDN (e.g., corn14.stanford.edu) with a local port tunnel to the TCP port that the server is listening on.

In Mac/Linux, setup the SSH tunnel in a terminal.

$ ssh jane@corn99.stanford.edu -L 9999:localhost:9999

substituting in the appropriate sunetid for jane, the appropriate FarmShare server hostname for corn99, and the appropriate TCP port for 9999. Windows users can setup an SSH local tunnel using menu options in putty

Once the ssh tunnel has been established, paste the https://localhos... URL provided from the jupyter_start script into a web browser to connect to the notebook home page. If this URL is misplaced, use tmux attach -d to bring up the virtual terminal and copy the URL again. Since the encryption keys were self-signed, the browser will warn about an insecure connection, but disregard these and continue/proceed past them.

Examples

Python 3

Jupyter python demo.png

MATLAB

Jupyter matlab demo.png

R

Jupyter r demo.png

SAS

Jupyter sas demo.png

Julia

Jupyter julia demo.png

Personal tools
Toolbox
LANGUAGES