OpenSlides docker-compose with traefik routing

Traefik Traefik v2 (latest)
1

Hey there.
I hope this is the right place to ask for help with my traefik v2 configuration.
My traefik config is very close to default and works with 2 simple services. But now I want to start an OpenSlides instance at openslides.mydomain.de and sadly the docker-compose.yml is rather complex and I can not make it work with traefik. It always ends up with " Error starting userland proxy: listen tcp 127.0.0.1:8000: bind: address already in use". So I guess it should be a simple routing issue that I can not figure out.
My OpenSlides docker-compose.yml is here:

# This configuration was created from a template file.  Before making changes,
# please make sure that you do not have a process in place that would override
# your changes in the future.  The accompanying .env file might be the correct
# place for customizations instead.
version: '3.4'

x-osserver:
  &default-osserver
  image: openslides/openslides-server:latest
  networks:
    - web
    - back
  restart: always
x-osserver-env: &default-osserver-env
    AMOUNT_REPLICAS: 1
    AUTOUPDATE_DELAY: 1
    DEMO_USERS: ""
    CONNECTION_POOL_LIMIT: 100
    DATABASE_HOST: "pgbouncer"
    DATABASE_PASSWORD: "openslides"
    DATABASE_PORT: 5432
    DATABASE_USER: "openslides"
    DEFAULT_FROM_EMAIL: "noreply@example.com"
    DJANGO_LOG_LEVEL: "INFO"
    EMAIL_HOST: "postfix"
    EMAIL_HOST_PASSWORD: ""
    EMAIL_HOST_USER: ""
    EMAIL_PORT: 25
    ENABLE_ELECTRONIC_VOTING: "True"
    ENABLE_SAML: "False"
    INSTANCE_DOMAIN: "http://openslides.mydomain.de:8000"
    JITSI_DOMAIN: ""
    JITSI_ROOM_PASSWORD: ""
    JITSI_ROOM_NAME: ""
    OPENSLIDES_LOG_LEVEL: "INFO"
    REDIS_CHANNLES_HOST: "redis-channels"
    REDIS_CHANNLES_PORT: 6379
    REDIS_HOST: "redis"
    REDIS_PORT: 6379
    REDIS_SLAVE_HOST: "redis-slave"
    REDIS_SLAVE_PORT: 6379
    REDIS_SLAVE_WAIT_TIMEOUT: 10000
    RESET_PASSWORD_VERBOSE_ERRORS: "False"
x-pgnode: &default-pgnode
  image: openslides/openslides-repmgr:latest
  networks:
    - dbnet
  labels:
    org.openslides.role: "postgres"
  restart: always
x-pgnode-env: &default-pgnode-env
  REPMGR_RECONNECT_ATTEMPTS: 30
  REPMGR_RECONNECT_INTERVAL: 10
  REPMGR_WAL_ARCHIVE: "on"

services:
  server:
    << : *default-osserver
    # Below is the default command.  You can uncomment it to override the
    # number of workers, for example:
    # command: "gunicorn -w 8 --preload -b 0.0.0.0:8000
    #   -k uvicorn.workers.UvicornWorker openslides.asgi:application"
    #
    # Uncomment the following line to use daphne instead of gunicorn:
    # command: "daphne -b 0.0.0.0 -p 8000 openslides.asgi:application"
    depends_on:

      - server-setup
    environment:
      << : *default-osserver-env
    secrets:
      - django

   server-setup:
    << : *default-osserver
    entrypoint: /usr/local/sbin/entrypoint-db-setup
    environment:
      << : *default-osserver-env
    secrets:
      - django
      - os_admin


    depends_on:
      - pgbouncer
      - redis
      - redis-slave
      - redis-channels

  client:
    image: openslides/openslides-client:latest
    restart: always
    depends_on:
      - server
    labels:
      - "traefik.enable=true"
      - "traefik.http.routers.wiki.rule=Host(`openslides.mydomain.de`)"
      - "traefik.http.routers.wiki.entrypoints=websecure"
      - "traefik.http.routers.wiki.tls.certresolver=myresolver"
    networks:
      - web
    ports:
      - "127.0.0.1:8000:80"

  pgnode1:
    << : *default-pgnode
    environment:
      << : *default-pgnode-env
      REPMGR_NODE_ID: 1
      REPMGR_PRIMARY: # This is the primary
    volumes:
      - "dbdata1:/var/lib/postgresql"

  pgbouncer:
    environment:
      - PG_NODE_LIST=pgnode1
    image: openslides/openslides-pgbouncer:latest
    restart: always
    networks:
      back:
        aliases:
          - db
          - postgres
      dbnet:
  postfix:
    image: openslides/openslides-postfix:latest
    restart: always
    environment:
      MYHOSTNAME: "localhost"
      RELAYHOST: "localhost"

    networks:
      - back
  redis:
    image: redis:alpine
    restart: always
    networks:
      back:
        aliases:
          - rediscache
  redis-slave:
    image: redis:alpine
    restart: always
    command: ["redis-server", "--save", "", "--slaveof", "redis", "6379"]
    depends_on:
      - redis
    networks:
      back:
        aliases:
          - rediscache-slave

  redis-channels:
    image: redis:alpine
    restart: always
    networks:
      back:
  media:
    image: openslides/openslides-media-service:latest
    environment:
      - CHECK_REQUEST_URL=server:8000/check-media/
    restart: always
    networks:
      web:
      back:
    # Override command to run more workers per task
    # command: ["gunicorn", "-w", "4", "--preload", "-b",
    #   "0.0.0.0:8000", "src.mediaserver:app"]

volumes:
  dbdata1:

networks:
  back:
  dbnet:
  web:
    external: true

secrets:
  django:
    file: ./secrets/django.env
  os_admin:
    file: ./secrets/adminsecret.env
2

Something else is listening on port 8000. You have shared no traefik config so it is difficult to say if this is in any way related to traefik.

Have you tried removing ports section, that would only be useful if you are bypassing traefik: