[Enhancement] Better docker compose deployment.

[ma-karai]

Enhancement for easy deployment

Firstly amazing development! Really great stuff

As far as I can see, to update my current installation I have to pull the repository again and rebuild the images and then redeploy, which I admit is quite easy. However, to set up automated image builds on hub.docker.com is also quite straightforward and would reduce the clutter on my (selfish) hard drives quite significantly. I am sure other agree as well.

It would be nice if immich_server and immich_microservices would be available on docker hub, and docker compose would be provided that I dont have to build.

Kind Regards
Mak

[alextran1502]

Hi @ma-karai

I am planning to work on this as well, haven't gotten a chance to get around it yet. I will update this issue once I have the auto-tagging docker image pipeline sorted out.

[ma-karai]

Outstanding :)

[alextran1502]

I've managed to whip up a Github workflow for pushing the images to Dockerhub.

You can try it out

• immich-microservices
• immich-server

[ma-karai]

Works like a charm!

file structure needed: (only the nginx config, and the env file)

and the simple docker compose file

version: "3.8"

services:
  immich_server:
    image: altran1502/immich-server:latest
    entrypoint: ["/bin/sh", "./entrypoint.sh"]
    volumes:
      - ${UPLOAD_LOCATION}:/usr/src/app/upload
    env_file:
      - .env
    environment:
      - NODE_ENV=production
    #ports:
     # - 3000:3000
    depends_on:
      - redis
      - database
    networks:
      - immich_network
    restart: unless-stopped

  immich_microservices:
    image: altran1502/immich-microservices:latest
    entrypoint: ["/bin/sh", "./entrypoint.sh"]
    volumes:
      - ${UPLOAD_LOCATION}:/usr/src/app/upload
    env_file:
      - .env
    environment:
      - NODE_ENV=production
    #ports:
     # - 3001:3001
    depends_on:
      - database
    networks:
      - immich_network
    restart: unless-stopped

  redis:
    container_name: immich_redis
    image: redis:6.2
    networks:
      - immich_network

  database:
    container_name: immich_postgres
    image: postgres:14
    env_file:
      - .env
    environment:
      POSTGRES_PASSWORD: ${DB_PASSWORD}
      POSTGRES_USER: ${DB_USERNAME}
      POSTGRES_DB: ${DB_DATABASE_NAME}
      PG_DATA: /var/lib/postgresql/data
    volumes:
      - ./pgdata:/var/lib/postgresql/data
    ports:
      - 5432:5432
    networks:
      - immich_network

  nginx:
    container_name: proxy_nginx
    image: nginx:latest
    volumes:
      - ./nginx-conf:/etc/nginx/conf.d
    ports:
      - 2283:80
      - 2284:443
    logging:
      driver: none
    networks:
      - immich_network
    depends_on:
      - immich_server

  # immich_tf_fastapi:
  #   container_name: immich_tf_fastapi
  #   image: tensor_flow_fastapi:1.0.0
  #   restart: always
  #   command: uvicorn app.main:app --proxy-headers --host 0.0.0.0 --port 8000 --reload
  #   build:
  #     context: ../machine_learning
  #     target: cpu
  #     dockerfile: ../machine_learning/Dockerfile
  #   volumes:
  #     - ../machine_learning/app:/code/app
  #     - ${UPLOAD_LOCATION}:/code/app/upload
  #   ports:
  #     - 2285:8000
  #   expose:
  #     - "8000"
  #   depends_on:
  #     - database
  #   networks:
  #     - immich_network

networks:
  immich_network:

[alextran1502]

I am glad it works for you! Please reopen this issue if you have any additional suggestions :) I will integrate this to the upstream setup