support-services

Support services

This folder contains a docker composition which can simulate an actual GeoNetwork 4 instance to be used for testing and development on GeoNetwork-UI.

The instance is already prepared with a few sample records and offers a reproducible development and testing environment.

Requirements

Docker Compose version 2+

Start

$ docker compose up -d

On the first run, the database will be populated with the settings and sample records.

A quicker startup can be obtained by only targeting the init service:

$ docker compose up init -d

To clear the volumes and let the database repopulate itself from scratch, run:

$ docker compose down -v

Specifying a different GeoNetwork version

By default, the version of GeoNetwork used as a backend is 4.2.5. You can specify another version like so:

$ GEONETWORK_VERSION=4.2.2 docker compose up -d

Note: If you want to run GN 4.2 on a recent ubuntu kernel or similar and you run into main ERROR Could not reconfigure JMX java.lang.NullPointerException" you may want to add some JAVA_OPTS to the GN service of the docker composition to bypass the issue during dev:

-Dcom.sun.management.jmxremote=false
-Dcom.sun.management.jmxremote.authenticate=false
-Dcom.sun.management.jmxremote.ssl=false
-XX:-UseContainerSupport

Access services

GeoNetwork can be accessed on http://localhost:8080/geonetwork.

Kibana can be used to inspect the ElasticSearch index and experiment with requests; it is accessible on http://localhost:5601.

Regenerate the initial database state

Running the following command will extract the current state of the database and store it as the new initial state:

$ docker compose exec database pg_dump -U geonetwork -d geonetwork -Fp > docker-entrypoint-initdb.d/dump.sql

Warning: Only run this command when running Geonetwork in version 4.2.2, as otherwise the DB will be migrated and e2e tests may fail.

Please also keep in mind that the initial state of the database should be as lightweight as possible and that changing it might break integration tests relying on it.

💡 When adding new datasets, if you want to minimize the modifications of e2e tests, manually adjust the creation/modification/id dates of the new datasets (either in the dump.sql file or via a database tool). Don't forget to reindex to take these changes into account on your local datahub. You will still have to update some e2e tests, those that rely on dataset counts.