This is a helm Chart for GeoNode, an open platform to manage and share geospatial information.
This will deploy the following containers:
- GeoNode, the main Web UI
- A Celery container to process background tasks
- GeoServer, the back-end managing storage of all geospatial data
- Their dependencies, i.e.:
- A PostGIS database (Postgres with the gis extension) to back GeoServer
- RabbitMQ, to back the Celery job queue
- A Kubernetes cluster
- Helm
The chart will automatically install required dependencies, i.e. a RabbitMQ broker and a Postgres database with postgis and postgis_topology extensions installed, and link them up.
- The built-in nginx container only supports http for now. It is expected that https support will typically be handled at the ingress level.
Refer to values.yaml for details of all the variables that can be overridden, and create your own overrides in a separate file, e.g. my-values.yaml.
In particular you will need to setup a number of secrets, as well as provide a valid public domain name that the application will be reachable on.
Then, install the helm release as usual.
# Clone the project
git clone https://github.com/one-acre-fund/geonode-k8s && cd geonode-k8s
# Install chart dependencies
helm dependency update deployment/geonode
# Override desired values in your own override file
vi my-values.yaml
# Install chart
helm install --values my-values.yaml my-geonode deployment/geonodeYou should see a bunch of new pods popping up:
$ kubectl get pods
NAME READY STATUS RESTARTS AGE
my-geonode-geonode-9c4667bbc-nnkmp 4/4 Running 0 7m8s
my-geonode-postgresql-0 1/1 Running 0 22h
my-geonode-rabbitmq-0 1/1 Running 0 22hBy default, no Ingress is installed, so you may have to use kubectl port-forward, or configure the main service to use NodePort or LoadBalancer in order to reach it from your machine.
- GeoNode docs
- geonode-docker the base Docker image