Introduction

This Ansible Role will install GeoNode and required dependencies onto an Ubuntu 14.04 (Trusty) host. It includes tasks for PostgreSQL+PostGIS, GeoServer, GeoNode, nginx, uswgi and also includes tasks for using AWS RDS databases. It is meant to be used with a GeoNode template project by following the workflow described here http://github.com/geonode/geonode-project

Tested with a known minimal working Ansible version of 1.9.3.

Installing roles from ansible galaxy

The ansible playbook that performs the provisioning depends on a few roles provided in the ansible galaxy. You can install these rolls with the following command in this directory:

ansible-galaxy install -r requirements.txt

Role Variables

• app_name - GeoNode project name (default: geonode)
• github_user - GitHub username that owns the project (default: GeoNode)
• code_repository - URL to the Code Repository (default: https://github.com/{{ github_user }}/{{ app_name }}.git)

The app_name variable will be used to set the database names and credentials. You can override this behavior with the following variables.

• db_data_instance - Database instance for spatial data (default: {{ app_name }})
• db_metadata_instance - Database instance for the application metadata (default: {{ app_name }}_app)
• db_password - Database password (default: {{ app_name }})
• db_user - Database user (default: {{ app_name }})

You can also change the war used to deploy geoserver with the following variable.

• geoserver_url - GeoServer war URL (default: http://build.geonode.org/geoserver/latest/geoserver.war)

Dataqs Processors

• roles/dataqs/templates/dataq_settings.py:
  • Change the 'DATAQS_APPS' setting to add/remove individual dataqs processors
  • Change the 'CELERYBEAT_SCHEDULE' setting to add/remove/modify scheduled dataqs celery tasks

Setting up a vagrant box

To configure a local development virtual machine, you will need to have virtualbox and vagrant installed. Note: You may need to change the IP configuration in the VagrantFile to a valid ip on the local network

$ vagrant up geoservices
$ vagrant ssh geoservices

Note: You may need to bring the vagrant box down and up for geonode to work.

$ vagrant halt
$ vagrant up

Deploying to ec2 (or other server)

Several variables have to be set correctly before deploying to a remote server. This can be achived by creating a custom inventory with the group [geoservices] and the host you will deploy too.

[geoservices]
XXX.XXX.XXX.XXX ansible_ssh_private_key_file=PATH_TO_PEM_FILE ansible_user=ubuntu deploy_user=ubuntu site_url=http://ec2-XXX-XXX-XXX-XXX.us-west-2.compute.amazonaws.com/ server_name=XXX-XXX-XXX-XXX-XXX.us-west-2.compute.amazonaws.com

Replace X's with the IP address of the remote server

• ansible_user - will be the user ansible SSHes in as
• deploy_user - will be the user used to deploy and install all the software (usually the same as ansible_user)
• ansible_ssh_private_key_file - the PEM file that corresponds to the ansible_user and provides passwordless ssh access
• site_url - the url of the website - used by geonode to identify its base URL
• server_name - the fully qualified domain name of the server

To deploy, run ansible-playbook -i /path/to/inventory playbook.yml From this directory.

Alternately, variables may be placed in a local variables file, e.g.:

/path/to/local_vars.yml

ansible_ssh_private_key_file: PATH_TO_PEM_FILE 
ansible_user: ubuntu 
deploy_user: ubuntu 
site_url: http://ec2-XXX-XXX-XXX-XXX.us-west-2.compute.amazonaws.com/ 
server_name: ec2-XXX-XXX-XXX-XXX.us-west-2.compute.amazonaws.com

With an inventory:

/path/to/inventory

[geoservices]
XXX.XXX.XXX.XXX 

To deploy, run:

ansible-playbook -i /path/to/inventory -e @/path/to/local_vars.yml playbook.yml