archivebox/docker-compose.yml

# Usage:
#     docker-compose run archivebox init --setup
#     docker-compose up
#     echo "https://example.com" | docker-compose run archivebox archivebox add
#     docker-compose run archivebox add --depth=1 https://example.com/some/feed.rss
#     docker-compose run archivebox config --set PUBLIC_INDEX=True
#     docker-compose run archivebox help
# Documentation:
#     https://github.com/ArchiveBox/ArchiveBox/wiki/Docker#docker-compose

version: '2.4'   # '3.9' or greater also works

services:
    archivebox:
        # build: .                              # for developers working on archivebox
        image: ${DOCKER_IMAGE:-archivebox/archivebox:master}
        command: server --quick-init 0.0.0.0:8000
        ports:
            - 8000:8000
        environment:
            - ALLOWED_HOSTS=*                   # add any config options you want as env vars
            - MEDIA_MAX_SIZE=750m
            # - SEARCH_BACKEND_ENGINE=sonic     # uncomment these if you enable sonic below
            # - SEARCH_BACKEND_HOST_NAME=sonic
            # - SEARCH_BACKEND_PASSWORD=SecretPassword
        # dns:                                  # uncomment to use pihole below for ad/tracker blocking during archiving
            # - pihole
        volumes:
            - ./data:/data
            # - ./archivebox:/app/archivebox    # for developers working on archivebox


    ### Optional Addons: tweak these examples as needed for your specific use case

    ### Example: To run the Sonic full-text search backend, first download the config file to sonic.cfg
    # curl -O https://raw.githubusercontent.com/ArchiveBox/ArchiveBox/master/etc/sonic.cfg
    # after starting, backfill any existing Snapshots into the index: docker-compose run archivebox update --index-only

    # sonic:
    #    image: valeriansaliou/sonic:v1.3.0
    #    expose:
    #        - 1491
    #    environment:
    #        - SEARCH_BACKEND_PASSWORD=SecretPassword
    #    volumes:
    #        - ./sonic.cfg:/etc/sonic.cfg:ro
    #        - ./data/sonic:/var/lib/sonic/store
    
    
    ### Example: To run pihole in order to block ad/tracker requests during archiving,
    # uncomment this block and set up pihole using its admin interface

    # pihole:
    #   image: pihole/pihole:latest
    #   ports:
    #     - 80:80       # uncomment to access the admin HTTP interface on http://localhost:80
    #   environment:
    #     WEBPASSWORD: 'set a secure password here or it will be random'
    #   volumes:
    #     - ./data/pihole:/etc/pihole
    #     - ./data/dnsmasq:/etc/dnsmasq.d


    ### Example: Run scheduled imports in a docker instead of using cron on the
    #     host machine, add tasks and see more info with archivebox schedule --help

    # scheduler:
    #    image: archivebox/archivebox:latest
    #    command: schedule --foreground --every=day --depth=1 'https://getpocket.com/users/USERNAME/feed/all'
    #    environment:
    #        - USE_COLOR=True
    #        - SHOW_PROGRESS=False
    #    volumes:
    #        - ./data:/data


    ### Example: Put Nginx in front of the ArchiveBox server for SSL termination

    # nginx:
    #     image: nginx:alpine
    #     ports:
    #         - 443:443
    #         - 80:80
    #     volumes:
    #         - ./etc/nginx/nginx.conf:/etc/nginx/nginx.conf
    #         - ./data:/var/www


    ### Example: run all your ArchiveBox traffic through a WireGuard VPN tunnel

    # wireguard:
    #   image: linuxserver/wireguard
    #   network_mode: 'service:archivebox'
    #   cap_add:
    #     - NET_ADMIN
    #     - SYS_MODULE
    #   sysctls:
    #     - net.ipv4.conf.all.rp_filter=2
    #     - net.ipv4.conf.all.src_valid_mark=1
    #   volumes:
    #     - /lib/modules:/lib/modules
    #     - ./wireguard.conf:/config/wg0.conf:ro


    ### Example: Run PYWB in parallel and auto-import WARCs from ArchiveBox

    # pywb:
    #     image: webrecorder/pywb:latest
    #     entrypoint: /bin/sh -c '(wb-manager init default || test $$? -eq 2) && wb-manager add default /archivebox/archive/*/warc/*.warc.gz; wayback;'
    #     environment:
    #         - INIT_COLLECTION=archivebox
    #     ports:
    #         - 8080:8080
    #     volumes:
    #         - ./data:/archivebox
    #         - ./data/wayback:/webarchive