imaging-plaza

This repository contains an script and docker-compose.yml able to deploy and run all the services needed for an imaging plaza cold start.

1. GraphDB
2. Gimie
3. SHACL validator
4. FAIR Indicator
5. Frontend

The following services still rely in third-party providers:

• Firebase Authentication
• Firebase Database

How to run it?

1. First add the credentials in ./config/.env.local file.
2. Uncompress the initial database (db.zip file not shared in this repo) in ./db.
   • Alternatively you can create an empty directory and initiate graphDB manually.
3. Execute the following docker compose commands

docker compose config
docker compose build
docker compose up

4. Access to the page in http://0.0.0.0:3000

Other services deployed are:

• GraphDB Workbench http://0.0.0.0:7515
• Fair Indicator GUI http://0.0.0.0:8502
• SHACL Validator GUI http://0.0.0.0:8501

How to run the developer mode?

Note

This repository makes use of submodules. Therefore, when cloning it you need to include them.

git clone --recurse-submodules https://github.com/sdsc-ordes/imaging-plaza

In development mode you can make use of the submodules to change to the development branch you would like to test.

docker compose -f docker-compose.dev.yml config
docker compose -f docker-compose.dev.yml build
docker compose -f docker-compose.dev.yml up

In order to start with a database please add the db.zip to db. GraphDB will take the file, uncompress it and then load the graph.

Also the submodules app folders are mounted as volumes in the /app folders in docker. This is optional, uncomment the volumes lines if you want to activate this option. Be aware that developing in windows can lead to issues due to different end of the line encodings.

For instace:

volumes:
    - "./gimie-api/app:/app
...

Finally, the entrypoint are replaced by bash so the services are required to be initialized manually. By default only the frontend is in development mode.

• gimie-api: uvicorn app.main:app --host 0.0.0.0 --port 15400
• shacl-validator-api: bash /app/entrypoint.sh
• fair-indicator-api: bash /app/imaging_plaza_fair_indicator_api/entrypoint.sh
• imaging-plaza-frontend: sh entrypoint.sh

If you want to disable this for some services, please comment the following lines in the docker-compose.dev.yml:

# entrypoint: bash 
# stdin_open: true
# tty: true

Once you have the different services up and running you can open with vscode (or any other IDE) the submodule folder and modify the different files. In the case of imaging-plaza-frontend those change will be applied automatically. If you want this, please uncomment the docker compose volumes.

Documentation

How to test a remote microservice?

Docker compose use internal alias for the services. The name of the container plus the port can be used in order to communicate among them. If you want to test a deployed server, please change the host in ./config/.env.local by the IP where the services is deployed:

For instance, if we want to connect to a remote graphDB deployed in REMOTE_IP you can do:

IP_DB_URL=http://REMOTE_IP:7200
IP_GIMIE_URL=http://gimie-api:15400/gimie/jsonld
IP_INFERENCE_URL=http://shacl-validator-api:15400/inference-jsonld?jsonInput=true&jsonOutput=true
IP_FAIR_LEVEL_INDICATOR=http://fair-indicator-api:15400/indicate