Supplying an Index Mapping
Now that installation is complete, we need to configure and supply an index mapping to Elasticsearch. This index mapping will complete Elasticsearch's setup and link the Index to the Arranger Server.
Using the Default Index Mapping
The default index mapping, called file_centric_1.0, is a sample mapping used for cancer genomics. It represents the structure of genomic file metadata that can be searched using Arranger. In Arranger, the sample mapping is configured in the index_config.json file, and can be viewed from the GitHub repository here.
To set up Arranger with this example mapping, run the following make command:
make seed-es
We recommend using Elasticvue to facilitate adding the Elasticsearch index. Use the following default information for connecting the index.
- Username: elastic
- Password: myelasticpassword
- URI: http://localhost:9200
Restart the Server
Once the Elasticsearch instance is connected you will need to restart Arranger server
You should now be able to interact with the Elasticsearch Index through localhost:5050/graphql
Supplying a Custom Index Mapping
The index mapping you create will reflect your own unique data model. You can use the sample default mapping described earlier as a guide, but the best resource for creating your own index mapping is the guidelines and best practices from Elastic.
Once you have drafted your index mapping, copy and paste it into the index_config.json file found here.
Naming your Index Mapping
You can rename the alias, but for ease-of-use it may be simpler to leave it as file_centric_1.0. If you change the alias, update the Make file variable ES_INDEX accordingly. You will also need to modify the make seed-es script, load-es-data.sh, to reference your updated alias name.
To update Arranger with your index mapping, restart Arranger Server and re-run the ES initialization script:
make seed-es
In the next section we will cover how to configure the search API and Interactive UI components using Arrangers metadata files.
Configuration