How to pass spring.datasource.data from docker as full path?

This is my docker-compose.yml file.

version: '3'
services:
  backend-service:
    build: ./backend-service
    volumes:
      - ./backend-service/database:/usr/lib/h2
    ports:
      - 8080:8080
    environment:
      - SPRING_DATASOURCE_URL=jdbc:h2:file:/usr/lib/h2/${DB}
      - SPRING_DATASOURCE_DRIVERCLASSNAME=org.h2.Driver
      - SPRING_DATASOURCE_USERNAME=SA
      - SPRING_DATASOURCE_PASSWORD=
      - SPRING_JPA_HIBERNATE_DDL-AUTO=update
      - SPRING_DATASOURCE_INITIALIZATION-MODE=always
      - SPRING_DATASOURCE_DATA=${DATA}
  frontend-service:
    build: ./frontend-service
    ports:
      - 3000:80
    depends_on:
      - backend-service

I want to pass a full system path to sql file like /usr/tmp/someSql.sql that should be executed when server is starting. I could use default data.sql but problem is that I will have multiple servers that will have different initial/update data so I don't want to put multiple files containing business data inside jar just to make it visible for spring boot and docker.

But even then if I put all the files inside resources folder and I have them in jar and use someData.sql as value it does not see that file during startup.

I also don't want to create endpoint for user to insert their own sql's for obvious reasons.

I already tried file:/usr/tmp/someSql.sql but I still get an error that file could not be found.

Okay a little late answer but I hope this will help someone in the future.

TL:DR version: I needed to add volume for scripts folder so my docker-compose.yml was extended by this:

volumes:
  ...
  - ./_init_scripts:/usr/init-data

And then point to it using SPRING_DATASOURCE_DATA:

  - SPRING_DATASOURCE_DATA=file:/usr/init-data/${DATA}

I am using .env file to store my configuration but it should work as long as scripts are present on docker enviroment and path is valid. eg.

  - SPRING_DATASOURCE_DATA=file:/usr/some/path/to/file/mySqlFile.sql

---

This was **TL:DR version ** now full story if someone would like to analyse it a little better:

I have 'docker project' folder structure like this:

DockerProjectRoot
  \_ _databases
   |  \_ myDatabase.mv.db
   |
   |_ _enviroments
   |  \_ env.myenv
   |
   |_ _init_scripts
   |  \_ someData.sql
   |
   |_ backend-service
   |  |_ ...
   |  \_ Dockerfile
   |
   |_ frontend-service
   |  |_ ...
   |  \_ Dockerfile
   |
   |_ docker-compose.yml

Directories:

• _databases - directory for runtime databases
• _enviroments - directory for storing .env files so I can start different enviroments with simple docker-compose --env-file ._enviroments/.env.myenv up command and change only .env file
• _init_scripts - directory where I would place data.sql files that need to be executed on different enviroments while creating new databases
• backend-service - backend app
• frontend-service - frontend app

contents of env.myenv are:

DB=myDatabase
DATA=someData.sql

And contents of docker-compose.yml

version: '3'
services:
  backend-service:
    build: ./backend-service
    volumes:
      - ./_databases:/usr/lib/h2
      - ./_init_scripts:/usr/init-data
    ports:
      - 8080:8080
    environment:
      - SPRING_DATASOURCE_URL=jdbc:h2:file:/usr/lib/h2/${DB}
      - SPRING_DATASOURCE_DRIVERCLASSNAME=org.h2.Driver
      - SPRING_DATASOURCE_USERNAME=SA
      - SPRING_DATASOURCE_PASSWORD=
      - SPRING_JPA_HIBERNATE_DDL-AUTO=update
      - SPRING_DATASOURCE_INITIALIZATION-MODE=always
      - SPRING_DATASOURCE_DATA=file:/usr/init-data/${DATA}
  frontend-service:
    build: ./frontend-service
    ports:
      - 3000:80
    depends_on:
      - backend-service

So when running I was running docker-compose --env-file ._enviroments/.env.myenv up it would use enviromental variables: DB and DATA to point to database location and initial scritps. So my enviroment dynamic valuse would be:

• SPRING_DATASOURCE_DATA=file:/usr/init-data/someData.sql - thing to note that it is full path inside docker container and it's connected to _init_scripts folder inside my docker project folder!
• SPRING_DATASOURCE_URL=jdbc:h2:file:/usr/lib/h2/myDatabase - again since docker image is connected via 'volume' this means myDatabse.mv.db will be updated between enviroment restarts

So in the end if I wanted to add new init-scripts I only need to add new .sql files to _init_scripts folder and update or create new .env file.