deploy-to-aws.mo

Owner: Xavier Lefèvre

Creation date: November, 13th 2017

Update date: November, 29th 2017

Context

• You want to host your back-end with Amazon Web Services

Prerequisites

• Have python3 and pip3 installed

• Have a Django project, see this if you want to start one

• Have an AWS account ready to pay services: https://portal.aws.amazon.com/billing/signup#/start

Steps

Create on AWS website an Identity and Access Management for the project (~5min)

• Find IAM from AWS services

• Go to the User tab

• "Add user"

• Give it a name and all access types

• Attach the below "existing policies directly"

  • AmazonRDSFullAccess

  • AmazonEC2FullAccess

  • AmazonElastiCacheFullAccess

  • AmazonS3FullAccess

  • AWSElasticBeanstalkFullAccess

• Retrieve and store somewhere safe the IAM user access and security key

Set-up your Elastic Beanstalk (EB) environment (~10min)

• Install EB CLI:

  pip3 install awsebcli

• Set up your EB environment, in your project root directory:

  eb init

  • Select your server location, close to where your customers will be:

    Select a default region
    1) us-east-1 : US East (N. Virginia)
    2) us-west-1 : US West (N. California)
    3) us-west-2 : US West (Oregon)
    4) eu-west-1 : EU (Ireland)
    5) eu-central-1 : EU (Frankfurt)
    6) ap-southeast-1 : Asia Pacific (Singapore)
    7) ap-southeast-2 : Asia Pacific (Sydney)
    8) ap-northeast-1 : Asia Pacific (Tokyo)
    9) sa-east-1 : South America (Sao Paulo)
    10) cn-north-1 : China (Beijing)
    (default is 3):

  • Connect to your IAM user

  • Give a general name to your instance, generally the name of the project/company, example: name_of_project-staging

  • Precise that you are using Python 3+

  • Say no to "CodeCommit"

  • Generate a new SSH Keypair that you will share with your teammates

    Check: EB should have generated a ".elasticbeanstalk" folder in your directory with your configuration

• You can commit finally your setup

  git add .
  git commit -m "Created EB project instance"

Configure your project for EB (~15min)

• We use cookiecutter as our base Django project at BAM, so we will use their configuration for EB:

  • Create a ".ebextensions" in your project root directory

  • Copy the EB configuration from cookie cutter repo

    • In "10_packages.config":

      packages:
        yum:
          git: []
          postgresql95-devel: []
          libjpeg-turbo-devel: []
          libffi: []
          libffi-devel: []

    • Copy "20_elasticcache.config", "30_options.config", "50_apache.config" and "enable_mod_deflate.conf" as is

    • Copy "40_python.config" but change the last two lines with:

      "aws:elasticbeanstalk:container:python:staticfiles":
        "/static/": "/staticfiles/"

• To connect to RDS, add a database configuration to your Django config in "base.py" or "production.py":

  # DATABASE CONFIGURATION
  # ------------------------------------------------------------------------------
  
  if 'RDS_DB_NAME' in os.environ:
    DATABASES = {
        'default': {
            'ENGINE': 'django.db.backends.postgresql_psycopg2',
            'NAME': env('RDS_DB_NAME'),
            'USER': env('RDS_USERNAME'),
            'PASSWORD': env('RDS_PASSWORD'),
            'HOST': env('RDS_HOSTNAME'),
            'PORT': env('RDS_PORT'),
        }
    }
  else:
      DATABASES['default'] = env.db('DATABASE_URL')

• To serve your statics from AWS, update also the Django config in "base.py":

  # STATIC FILE CONFIGURATION
  # ------------------------------------------------------------------------------
  
  STATIC_ROOT = 'staticfiles'
  STATIC_URL = '/staticfiles/'

• You can commit your setup

  git add .
  git commit -m "Added python, django and database config and requirements files"

Deploying the application (~45min)

• For the first time you will create the instance, hence all the necessary servers:

  eb create --scale 1 -db -db.engine postgres -db.i db.t2.micro

  • Name your environment, like: project-name-staging

  • Select the default CNAME

  • Select the classic load balancer option

  • Choose a username and password to connect to the DB

• EB will now generate all your servers and provision your config then launch your project

• But you might still have errors eb logs, you are missing environment variables that you can directly add to AWS website

  • On AWS website, on the top right, select the right server location (ex: London)

  • Go to Elastic Beanstalk

  • Enter your new application

  • Go to configuration

  • Modify the software configuration with variables such as:

    DJANGO_ACCOUNT_ALLOW_REGISTRATION, DJANGO_ADMIN_URL, DJANGO_ALLOWED_HOSTS, DJANGO_AWS_ACCESS_KEY_ID, DJANGO_AWS_SECRET_ACCESS_KEY, DJANGO_MAILGUN_API_KEY, DJANGO_SECRET_KEY, DJANGO_SECURE_SSL_REDIRECT, DJANGO_SENTRY_DSN, DJANGO_SERVER_EMAIL, DJANGO_SETTINGS_MODULE, MAILGUN_SENDER_DOMAIN, REDIS_ENDPOINT_ADDRESS, REDIS_PORT

    Check 1: You can verify on the AWS website that each server piece has been created (EB, RDS, S3, EC2) Check 2: You can access the aws.ip/admin

Create a superadmin (~10min)

Now that you have deployed once, on your host you should have the credentials for AWS (cat ~/.aws/config).

To create a super user you will have to connect directly to the machine in order to run ./manage.py createsuperuser.

• To ssh to your EC2 machine run eb ssh [PUT YOUR INSTANCE NAME IF NOT BY DEFAULT]

• Then on the machine go to the root of the project:

  cd /opt/python/current

• Add the python environment:

  source ./env

• Run the createuser command:

  cd /app
  ./manage.py createsuperuser

  Check: You can access the aws.ip/admin and go through now with your newly created superadmin

Debug

eb logs

Deployment

eb deploy

Sources

• Cookie-cutter Django repository
• Step by step tutorial to deploy Django to AWS EB
• Another tutorial to deploy Django to AWS EB