I would like to be able to deploy my app in a pre-prod
environment for integration testing using a Docker volume that will expose an instance of PostgreSQL. I'm using Scala v2.12.8 and Play v2.7.
Looking at the environment settings of the SBT native packager it seems possible to define dockerExposedVolumes
in order to attach a DB.
Using a normal Docker compose file I would do something like that:
version: "3"
services:
db:
image: postgres
environment:
- POSTGRES_USER=postgres
- POSTGRES_PASSWORD=postgress
- POSTGRES_DB=postgres
ports:
- "5433:5432"
volumes:
- pgdata:/var/lib/postgresql/data
networks:
- suruse
volumes:
pgdata:
This configuration has been taken from this SO answer.
I tried searching for config examples but I didn't find anything useful so far. Now I'm wondering how I should define a new docker volume and then expose it to the Docker image created by SBT exactly?
The final version. I've fully tested it and it works exposing the DB on the TCP port 5433.
# https://docs.docker.com/samples/library/postgres/
version: "3"
services:
app-pgsql:
image: postgres:9.6
restart: always
environment:
- POSTGRES_USER=postgres
- POSTGRES_PASSWORD=yourPasswordHere
- POSTGRES_DB=yourDatabaseNameHere
- POSTGRES_INITDB_ARGS="--encoding=UTF8"
ports:
- "5433:5432"
volumes:
- pgdata:/var/lib/postgresql/data
volumes:
pgdata:
driver: local
Launch the docker compose using sbt dockerComposeUp -useStaticPorts
and then check if the containers have been actually exposed using docker ps -a
. Also, check the log files using the command provided by dockerComposeUp
or dockerComposeInstances
.
There is a sbt Plugin that helps you to achieve this:
With that you can add your database to a docker compose file and you can run everything within sbt.
This is a Docker standard. Here is an explaination how to do it for Postgres:
[run_postgresql_docker_compose][2]
The docker-compose.yml
from that example:
version: '3'
services:
mydb:
image: postgres
volumes:
- db-data:/var/lib/postgresql/data
ports:
- 5432:5432/tc
volumes:
db-data:
driver: local
As this is a standard way of Docker you will find more examples.