EEA Search Node.js application + Docker env

Development

1. Change code
2. If you:
3. Want to change something in eea.searchserver.js: ./build_dev.sh $PATH_TO_EEA_SEARCHSERVER_JS_REPO
4. If you changed Dockerfile or app/package.json: docker build -t eeacms/eeasearch:dev . and also make the changes in Dockerfile.dev.
5. If you changed only something in './app' : go to step 3.
6. If step 3 is still running and last build was made with ./build_dev.sh: go to step 4 because the server will automatically restart inside the container
7. Note: Dockerfile.dev builds only with ./build_dev.sh because it temporarely adds eea.searchserver.js in the current directory.
8. docker-compose up
9. Go to http://localhost:3000
10. Before pushing to master make sure that ./test.sh passes.

If you want to check that the image is built correctly, please comment out the volumes entry in docker-compose.yml

No Docker Development

1. cd app
2. npm install
3. Set elastic_* env variables
4. ./app.js runserver
5. Go to http://localhost:3000

Production

Deploy via Rancher

The application is available in the EEA Rancher Catalog.

Other

1. docker pull eeacms/eeasearch
2. Use any orchestration solution and make sure that the container can ping the Elastic backend
3. Pass commands and env variables to the image as explained in the Docker Usage section

Docker Usage

Basic usage of the image is given by the following pattern:

docker run -e elastic_host=$YOUR_ELASTIC_HOST run eeacms/eeasearch $command

To see the available commands run:

docker run -e elastic_host=$YOUR_ELASTIC_HOST run eeacms/eeasearch help

Available commands:

• create_index: create the index on elastic_host and start harvesting
• reindex : recrate the index on the elastic_host and start harvesting
• sync_index: sync the resources
• remove_data: remove the ES index of this app
• remove_river: stop the harvesting process by removing the river

Environment variables:

• elastic_index: the elasticsearch index id where to store the data.
  • note the elastic index in app/config/settings.json takes precendence, so delete if you want the environment variable to work.
• elastic_host, elastic_port, elastic_path: the elasticsearch endpoint parameters the elasticsearch endpoint will be interpreted as such: http://$elastic_host:$elastic_port$elastic_path
• by default, elastic_port is 9200
• by default, elastic_path is /
• NODE_ENV: can be either dev or production
• dev will do a more verbose logging
• production will log only erros in an APACHE format
• Assuming that the app will be proxied in production, the proxy app should contain the access logs
• SYNC_CRONTAB: A valid crontab line (e.g. * * * * *) for scheduling sync jobs. If not set, the app will never schedule sync jobs. In the image this variable is not set.

Note that any index creation commands require rights on the set elastic_backend

For more information about the settings.yml file, please see https://github.com/eea/eea.searchserver.js/blob/master/README.md