ArchiveBox

version: "3.3"
services:
    archivebox:
        image: archivebox/archivebox:latest
        ports:
            - 8000:8000
        volumes:
            - ./data:/data
        environment:
            # - ADMIN_USERNAME=admin            # create an admin user on first run with the given user/pass combo
            # - ADMIN_PASSWORD=SomeSecretPassword
            - ALLOWED_HOSTS=*                   # restrict this to only accept incoming traffic via specific domain name
            - PUBLIC_INDEX=True                 # set to False to prevent anonymous users from viewing snapshot list
            - PUBLIC_SNAPSHOTS=True             # set to False to prevent anonymous users from viewing snapshot content
            - PUBLIC_ADD_VIEW=False             # set to True to allow anonymous users to submit new URLs to archive
            - SEARCH_BACKEND_ENGINE=sonic       # tells ArchiveBox to use sonic container below for fast full-text search
            - SEARCH_BACKEND_HOST_NAME=sonic
            - SEARCH_BACKEND_PASSWORD=SomeSecretPassword
            # - PUID=911                        # set to your host user's UID & GID if you encounter permissions issues
            # - PGID=911                        # UID/GIDs <500 may clash with existing users and are not recommended
            # - MEDIA_MAX_SIZE=750m             # increase this filesize limit to allow archiving larger audio/video files
            # - TIMEOUT=60                      # increase this number to 120+ seconds if you see many slow downloads timing out
            # - CHECK_SSL_VALIDITY=True         # set to False to disable strict SSL checking (allows saving URLs w/ broken certs)
            # - SAVE_ARCHIVE_DOT_ORG=True       # set to False to disable submitting all URLs to Archive.org when archiving
            # ...
            # add further configuration options from archivebox/config.py as needed (to apply them only to this container)
            # or set using `docker compose run archivebox config --set SOME_KEY=someval` (to persist config across all containers)
        # For ad-blocking during archiving, uncomment this section and pihole service section below
        # networks:
        #   - dns
        # dns:
        #   - 172.20.0.53
 
    archivebox_scheduler:
        image: archivebox/archivebox:latest
        command: schedule --foreground --update --every=day
        environment:
            - TIMEOUT=120                       # use a higher timeout than the main container to give slow tasks more time when retrying
            # - PUID=502                        # set to your host user's UID & GID if you encounter permissions issues
            # - PGID=20
        volumes:
            - ./data:/data
        # cpus: 2                               # uncomment / edit these values to limit scheduler container resource consumption
        # mem_limit: 2048m
        # restart: always
 
    sonic:
        image: valeriansaliou/sonic:latest
        build:
            # custom build just auto-downloads archivebox's default sonic.cfg as a convenience
            # not needed after first run / if you have already have ./etc/sonic.cfg present
            dockerfile_inline: |
                FROM quay.io/curl/curl:latest AS config_downloader
                RUN curl -fsSL 'https://raw.githubusercontent.com/ArchiveBox/ArchiveBox/stable/etc/sonic.cfg' > /tmp/sonic.cfg
                FROM valeriansaliou/sonic:latest
                COPY --from=config_downloader /tmp/sonic.cfg /etc/sonic.cfg
        expose:
            - 1491
        environment:
            - SEARCH_BACKEND_PASSWORD=SomeSecretPassword
        volumes:
            - ./sonic.cfg:/etc/sonic.cfg
            - ./data/sonic:/var/lib/sonic/store
 
    novnc:
        image: theasp/novnc:latest
        environment:
            - DISPLAY_WIDTH=1920
            - DISPLAY_HEIGHT=1080
            - RUN_XTERM=no
        ports:
            # to view/control ArchiveBox's browser, visit: http://127.0.0.1:8080/vnc.html
            # restricted to access from localhost by default because it has no authentication
            - 8080:8080