Basic Linux Packages¶

unzip – sudo apt-get install unzip pip – sudo apt-get install python-pip python-dev – sudo apt-get install python-dev

-> used by pandas installer

Installing¶

sudo apt-get update sudo apt-get install postgresql postgresql-contrib

you should see something like this:

Removing obsolete dictionary files:

• No PostgreSQL clusters exist; see “man pg_createcluster”

Processing triggers for ureadahead (0.100.0-16) ... Setting up postgresql-9.3 (9.3.11-0ubuntu0.14.04) ... Creating new cluster 9.3/main ...

config /etc/postgresql/9.3/main data /var/lib/postgresql/9.3/main locale en_US.UTF-8 port 5432

Now login to the postgres user with this command:

sudo -i -u postgres

then enter the psql shell by running:

psql postgres

now you should be able to create datases and roles.

Adding additional Linux Helper Packages¶

sudo apt-get install python-psycopg2 sudo apt-get install libpq-dev

Setting up Users and Creating the DB¶

In the settings.py file that will be used by the application to handle sensitive information, branding and the management of packages, we will have a database host and user cofiguration. So whatever you decide to call your database here, will be the information that is used by the application to access the Database.

1. Create a User:

WHile in the postgres shell run:

CREATE ROLE djangoapp WITH PASSWORD ‘mySecurePassword’ LOGIN SUPERUSER;

Make sure your password is secure and uses numbers, letters and special characaters

after exiting the psql shell with postgres user, Make sure this user was created by logging in like so :

pql postgres -U djangoapp

and follow the prompt for password.

If you see this error:

-> psql: FATAL: Peer authentication failed for user “djangoapp”

You likely will have to go to your pg_hba.conf file and make sure local requests are handled with “md5” and NOT peer authentication.

see more here: http://stackoverflow.com/questions/18664074/getting-error-peer-authentication-failed-for-user-postgres-when-trying-to-ge

you will want to reload postgres by running:

–> sudo /etc/init.d/postgresql reload

For more on postgres, installing and debugging see here: https://help.ubuntu.com/community/PostgreSQL

2. Create the database

CREATE DATABASE afg_eoc WITH OWNER djangoapp;

now make sure you can login to the database with:

psql afg_eoc -U djangoapp

NOTE: - the config for postgres is /etc/postgresql/9.3/main - the data for postgres is /var/lib/postgresql/9.3/main

3. create all of the necessary application files and diretories. Create App Directory and File System ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

If you look at fabfile.py at the root of the directory, it relies on a few directories to be in place in order to deploy and serve up the applicaion.

remote_work_path = ‘~/deploy/rhizome-work’ remote_backend_path = ‘/var/www/apps/rhizome/’ remote_frontend_path = ‘/var/www/apps/rhizome/webapp/public/static/’

Give ownership to the user that deploys your system.. so if your ssh config for the new remote server is

myUser

execute on your machine:

sudo chown -R myUser /var/www/apps/rhizome/

To make sure that you are able to create and access the above directores.

Finally create the “venv” path which is where we store all of the packages of application.

mkdir /var/www/apps/rhizome/venv

4. push up the code using the fabfile.

If you have the .ssh/config set up like above you can deploy by

1. entering the virtualenv with:

source/venv/bin/activate

2. fab -H mySshConfig deploy

Check for failiures.. the execution should tell you at the bottom exactly what line failed.. so ssh into the server, run that command and see if you can track down the issue.

When spinning up a new ubuntu instance nstall the following dependencies:

sudo apt-get update
sudo apt-get install apache2 --fix-missing
sudo apt-get install unzip --fix-missing
sudo apt-get install python-pip --fix-missing
sudo apt-get install python-pandas --fix-missing
sudo apt-get install python-dev
sudo apt-get install libpq-dev
sudo apt-get install postgresql-9.3
sudo apt-get install python-psycopg2
sudo apt-get install libapache2-mod-wsgi
sudo apt-get install apache2 apache2 apache2-mpm-prefork apache2-utils libexpat1 ssl-cert

To set up the database:

CREATE USER djangoapp WITH PASSWORD 'somepassword' SUPERUSER LOGIN;

Then create the directory for the rhizome django app, and static files:

sudo mkdir /var/www/apps/
sudo mkdir /var/www/apps/rhizome/

You’ll need to configure Apache and WSGI on whatever server you’re deploying to, and set up a PostgreSQL database for the application first. Make sure that Apache is configured to use the prefork MPM; the worker and event MPMs result in incorrect responses returned for requests when multiple requests are made to the server.

For more information on deploying applications, see the Django documentation.

You will also want to set up apache to server the files in the /static and /media directories.

Serving Static files¶

The frontend files can be served statically by any webserver. Make sure that the STATIC_URL setting in the backend’s settings.py is set to wherever you deploy the static files. ([Read more about static files in Django.][static-files])

static-files

We now gather all the files required for the apache server to send out to the users:

cd webapp
npm run package

This will create a new file under the dist/ folder rhizome.zip. Now we move this file to the server and unzip. into the /var/www/apps/rhizome/ folder. To install all of the required dependencies we run:

pip install -r requirements.txt


Creating a New Server on AWS UNICEF Account
-------------------------------------------

  - Go to AWS -- Click "Ec2"
  - Launch Instance
      -> Ubuntu Server 14.04 LTS (HVM), SSD Volume Type
      -> type m4.large
    click "Configure Instance Details" -- NOT review and launch
    use the default selected here.. and click add storage
  - Storage Screen - Select 64-g, only one volumne needed, next click "tag Instance"
  - Tag Instance * whatever you want *
  - Configure Security Group
      -> Choose "Rhizome Prod Secuity Group"
  - Click "review and launch"

At this point your server should spin up .. now lets set up an Elastic IP

Elastic I.P.
^^^^^^^^^^^

the point of an elastic I.P. is that you can assign mutliple machines to this address so that if you have to swap out the server, or spin up a new one, and it has a different i.p. you can simply associate the old instance to the elastic IP.

Once the server is "running" * ( see console.aws.amazon.com/ec2  ) go to "elastic IPS"
  - Allocate New Address -> this creates a publc I.P. that can be used via a domain name router like go datddy or network solutions.
  - Select the new instance by clicking the check box at the left
  - click actions >> asscoiate addres
      - choose the instance that you just created and click "associate"
  - For conveninve you will want to add the new "public IP" along wiht your ssh information to your ~/.ssh/config

  for instance

Host myServer
    User john
    HostName my.ip.adr.01
    IdentityFile ~/.ssh/some_key.pem

Once that it set up, you should be able to .ssh into the machine simply by running:

ssh myServer