Docker & Docker Compose

Grouparoo is easy to deploy with Docker and Docker Compose.

Example

An example project can be found at https://github.com/grouparoo/app-example

Getting Started

1. Create a new git repository
   • Do not include an environment file, .env. Your environment variables will be managed by Docker
   • Include a package.json file which contains your dependencies and commands per the deployment Guide
   • Include a .dockerignore file which excludes node_modules and any local config you might have (ie: .env)
   • Include a Dockerfile which is similar the to the below. We are building a Dockerfile which can be used in multiple modes, as a background task worker or web server, which are controlled by Environment Variables.

# Dockerfile
FROM node:12

WORKDIR /grouparoo

ENV NODE_ENV='production'
ENV PORT=3000
ENV WEB_URL=http://localhost:$PORT
ENV WEB_SERVER=true
ENV SERVER_TOKEN="change-me"
ENV SCHEDULER=true
ENV WORKERS=1
ENV REDIS_URL="redis://localhost:6379/0"
ENV DATABASE_URL="postgresql://localhost:5432/grouparoo_development"
ENV S3_ACCESS_KEY=""
ENV S3_SECRET_ACCESS_KEY=""
ENV S3_REGION=""
ENV S3_BUCKET=""

COPY . .
RUN npm install
RUN npm run prepare
RUN npm prune

WORKDIR /grouparoo/node_modules/@grouparoo/core
CMD ["./api/bin/start"]

EXPOSE $PORT/tcp

2. Build your docker image with docker build .
3. Push and run your docker image!

Docker Compose

As an example of how to orchestrate a more complex deployment, the following example docker-compose file is provided. This will create our needed networks (frontend and backend), our needed storage servers (Postgres and Redis), 2 types of grouparoo instances (web and backend).

This example requires a docker swarm cluster (or local instance). You can create one on your local machine with docker swarm init.

# docker-compose.yml
version: "3.1"
services:
  redis:
    image: redis
    restart: always
    networks:
      - grouparoo_backend

  db:
    image: postgres
    restart: always
    environment:
      POSTGRES_PASSWORD: password
      POSTGRES_DB: grouparoo_docker
    networks:
      - grouparoo_backend

  grouparoo-web:
    build: ./
    ports:
      - 3000:3000
    environment:
      PORT: 3000
      REDIS_URL: redis://redis:6379/0
      DATABASE_URL: postgresql://postgres:password@db:5432/grouparoo_docker
      WEB_SERVER: "true"
      SCHEDULER: "false"
      WORKERS: 0
    depends_on:
      - db
      - redis
    networks:
      - grouparoo_frontend
      - grouparoo_backend

  grouparoo-worker:
    build: ./
    environment:
      REDIS_URL: redis://redis:6379/0
      DATABASE_URL: postgresql://postgres:password@db:5432/grouparoo_docker
      WEB_SERVER: "false"
      SCHEDULER: "true"
      WORKERS: 10
    depends_on:
      - db
      - redis
    networks:
      - grouparoo_frontend
      - grouparoo_backend

networks:
  grouparoo_frontend:
    driver: overlay
  grouparoo_backend:
    driver: overlay

Notes

Remember, all Environment Variables can be changed from their defaults, including database information, PORT, etc.

Docker will regularly restart/rebuild your images. You may at times see "stuck" background task in the resque dashboard. Grouparoo will automatically rescue these jobs and retry them after an hour.

All the Environment Variables have defaults, but you are expected to customize them. You likely will want to create multiple runtime clusters, some workers and some web servers. This way you can scale each runtime mode separately as needed.

The example docker-compose.yml has no data persistence - DO NOT USE IN PRODUCTION!

The example docker-compose.yml has no load balancing between grouparoo-web instances - DO NOT USE IN PRODUCTION!

You can scale the number of web and worker processes independently. Depending on your workload, you may need more of one type of process than another.