Run Jupyter Notebook on Amazon-EC2

In this tutorial, i will share to you the quickest way to run ipython notebook using Amazon EC2 (Amazon Elastic Cloud Computing).

Create AWS account: https://aws.amazon.com/ and login to the console

Launch an EC2 instance

Name your EC2 instance as MyFirstInstance

Select an operating system: Ubuntu

Select instance type: t2.micro

Create a key pair: MyFirstInstance.pem and save the key pair in your pc

Highlight the instance and click on Security Groups at the left side of the console:

Click on Actions and Edit inbound rules:

Add rules for Custom TCP, SSH and HTTPS

Your AWS-EC2 instance is ready and the following steps will allow your to connect to this EC2 instance.

Using Putty for windows for SSH (remote shell) and SFTP (transfer files)

Install Putty for windows here: https://www.chiark.greenend.org.uk/~sgtatham/putty/latest.html

Convert MyFirstInstance.pem into MyFirstInstance.ppk by using puttygen.exe

Double click on puttygen.exe then File -> Load private key (MyFirstInstance.pem) -> Save private key -> use file name MyFirstInstance.ppk

For convenience, save MyFirstInstance.ppk on the same folder where you install Putty

Go back to EC2 console and copy the Public DNS of your instance

Connect to your EC2 instance via SSH to install Anaconda

Open putty.exe and go to Connection-->SSH->Auth

Load the private key (MyFirstInstace.ppk)

Click on Session then input info on host name: <copied from EC2 DNS> then save your session as AWS then click Open

Login as Ubuntu and you are now connected to the EC2 instance. The authentication key (.ppk) allows you to login to EC2 server without a password.

Visit Anaconda and copy link the installer for Linux

https://www.continuum.io/downloads

Download Anaconda to your EC2 server:

wget https://repo.continuum.io/archive/Anaconda3-4.2.0-Linux-x86_64.sh

Install Anaconda

bash Anaconda3-4.2.0-Linux-x86_64.sh

Once Anaconda is installed; create config file for jupyter

cd /home/ubuntu/anaconda3/bin

jupyter notebook --generate-config

cd ~/.jupyter/

vi jupyter_notebook_config.py

add below lines in notebook_config.py

c.NotebookApp.ip = '*' c.NotebookApp.open_browser = False #do not open browser c.NotebookApp.port = 8888 #fix port at 8888

Create a new screen and start jupyter notebook

screen

This will open a new terminal then run the jupyter notebook

jupyter notebook

Now your notebook is running in AWS-EC2

To exit from this screen, Ctrl – a d

To re-open the previous screen type, screen -r

The jupyther notebook is now running in your AWS-EC2 virtual machine and can be accessed from your windows browser

Open firefox or chrome browser and put the ECS DNS and using port 8888

http://<DNS>.us-east-2.compute.amazonaws.com:8888/tree

Set-up a FTP user for ftp (to upload files)

Create a userid in EC2. Sudo allows you to run a command using a super user with proper security privileges

sudo passwd www-data

Set-up the login shell

sudo vim /etc/passwd

Change www-data:x:33:33:www-data:/var/www:/usr/sbin/nologin to /bin/bash

www-data:x:33:33:www-data:/var/www:/bin/bash

Test the connectivity via FTP

Go to your Putty folder and open sftp.exe

Type open www-data@<EC2 DNS> then enter password of www-data

Now you can upload files to your EC2 instance

End of tutorial