Skip to content

Latest commit

 

History

History
44 lines (32 loc) · 1.57 KB

README_DOCKER.md

File metadata and controls

44 lines (32 loc) · 1.57 KB
Raw

Docker

Although we have previously covered how to run each platform component on its own, there may be instances where you want to test the platform in an isolated environment. Since each DL4SE component equates to a container, we make use of Docker Compose. Our platform is compartmentalized into five services:

  • dl4se-db: Responsible for hosting our platform database;
  • dl4se-liquibase: Used to automatically perform database migrations on startup;
  • dl4se-crawler: Host to the crawler application;
  • dl4se-server: Host to the back-end server;
  • dl4se-website: Responsible for distributing the website resources to clients.

Setup

Dockerizing the platform requires first creating the .env file that docker-compose will use to create and start all the defined services. This file essentially contains the environment variables that we have defined beforehand. Although you could manually create this file with all the settings, if you have set up the environment variables, then you can quickly generate this file using the deployment/envgen.sh script. The final file should look something like this:

COMPOSE_PROJECT_NAME=dl4se

DATABASE_NAME=dl4se
DATABASE_USER=dl4se_admin
DATABASE_PASS=Lugano2022
DATABASE_PORT=5432

# ...

Deployment

After that, creating the containers and starting all the services is as simple as running the following:

docker-compose -f deployment/docker-compose.yml up -d

Shutting down the services and removing the created containers:

docker-compose -f deployment/docker-compose.yml down