This is the repository for the poms project at Kings Digital Lab
This project uses the technologies outlined in our Technology Stack and is configured to use Vagrant for local development and Fabric to now use Docker for deployment ( see below).
- nginx-proxy: This is the primary entry point for the stack, running on 80. It automatically builds a proxy to other containers.
- django 3.2: The main container for the project (see more detailed description below.)
- nginx: This is the static data container, serving Django's static content.
- db: The database container for Django above, running a legacy version of MySQL (5.7).
- elasticsearch 7.10: The indexing container, used by Haystack 3.2.1. (Pre-migration, Haystack 2 was using Solr 6.)
- geoserver: The original VM deployment shared a geoserver with other projects. It has now been dockerized and moved here as a standalone resource.
- postgis: The backing database container for the geoserver above.
- rdf: This container is an encapsulation of a Tomcat server running John Bradley's modified RDF4J to provide the dataset as linked open data. For more information, see the at documentation
The compose file will look for deployment variables in a compose/.env file. Below is a sample file:
#Django
DJANGO_READ_DOT_ENV_FILE=True
DJANGO_ALLOWED_HOSTS=
DJANGO_SECRET_KEY=
DJANGO_DEBUG=False
# Elasticsearch
# ------------------------------------------------------------------------------
discovery.type=single-node
# MySQL
# ------------------------------------------------------------------------------
DATABASE_URL=
MYSQL_ROOT_PASSWORD=
MYSQL_DATABASE=
MYSQL_USER=
MYSQL_PASSWORD=
MYSQL_HOST=db
# PostGIS
# ------------------------------------------------------------------------------
POSTGRES_USER=
POSTGRES_PASSWORD=
# GEOSERVER
geoserver_admin_user=
geoserver_admin_password=
Fill in the database credentials and Django variables. If deploying via a CI pipeling such as Gitlab, this file will need to be included in its variables (in the KDL setup, we encode this in base64 and add it to the CI/CD variables in the repository settings.)
- Enter the project directory:
cd poms-django - Build and run the docker containers
docker compose -f compose/docker-compose.yml up -d --build - Copy mysql data into the db container and ingest it via the command line if necessary.
- Haystack indexes are not buit during the build process. To build the Haystack search indexes, first log into the django container
docker compose -f compose/docker-compose.yml exec django bash. Then run the update_index management command:python manage.py update_index
You can access the site locally at http://localhost