Cloud 101 For AWS

Doing computations in the cloud

Overview

Teaching: 0 min
Exercises: 40 min
Questions
  • What can we use a cloud machine for?

Objectives
  • Learn how to install miniconda and related packages on your cloud machine

  • Learn how to enable the Jupyter Notebook server on your cloud machine and access it from a browser

  • Learn how to manipulate data stored on an s3 bucket

How to do computations on your cloud machine

Now, we have data inside our cloud machine. Let’s see how we can do some computing with this data.

To install Python-related software, we’ll make sure that our machine has the packages required for us to work with.

Step 1. Install miniconda wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh

bash Miniconda3-latest-Linux-x86_64.sh

Press spacebar to scroll through the manual. Answer yes to the prompts. After miniconda is installed, we want to make sure our conda path is updated. This can be done using:

source ~/.bashrc

Step 2. Create a conda environment and all packages within the environment

conda create -n esip python=3.7

conda activate esip

conda install jupyter notebook

conda install matplotlib rasterio boto3

It will take a while to get through these steps

Step 3. Get Jupyter Notebook going

jupyter notebook --ip="0.0.0.0"

You will see something pop up that looks like this:

To access the notebook, open this file in a browser:
        file:///home/ubuntu/.local/share/jupyter/runtime/nbserver-2063-open.html
    Or copy and paste one of these URLs:
        http://(ip-172-31-36-8 or 127.0.0.1):8888/?token=90dcf8abfc8e205ff0bdac22c260bce9cb01e2f34d6545a8
  

Copy the URL to the address bar of your web browser. Replace “localhost” with the Public IP of your Virtual Machine.
We are using 0.0.0.0 here as the IP to enable ALL IP addresses to reach the notebook server.

Step 4. Let’s do some fun stuff with Jupyter Notebooks Once you have Jupyter Notebook going, we will follow the steps outlined in this notebook: https://github.com/amanda-tan/cloud101_aws_esip/blob/gh-pages/landsat_s3_ndvi.ipynb

We will launch a new Python 3 notebook and do some plotting with rasterio

Key Points