docker-compose with-nginx-proxy example with self-signed ssl crashes with no debug logs

[LeaderOne]

When running the compose.yaml file from .examples/.examples/docker-compose/with-nginx-proxy/mariadb/apache with a self-signed certificate, nginx will crash with little to no explanation of what is happening. The example, running with a MYSQL_PASSWORD set and the db.env file appropriately updated, shows log entries like this, but does not explain why it is shutting down:

nginx.1     | 2025/02/08 14:22:09 [notice] 17#17: using the "epoll" event method
nginx.1     | 2025/02/08 14:22:09 [notice] 17#17: nginx/1.27.3
nginx.1     | 2025/02/08 14:22:09 [notice] 17#17: built by gcc 13.2.1 20240309 (Alpine 13.2.1_git20240309) 
nginx.1     | 2025/02/08 14:22:09 [notice] 17#17: OS: Linux 6.1.118-Unraid
nginx.1     | 2025/02/08 14:22:09 [notice] 17#17: getrlimit(RLIMIT_NOFILE): 40960:40960
nginx.1     | 2025/02/08 14:22:09 [notice] 17#17: start worker processes
nginx.1     | 2025/02/08 14:22:09 [notice] 17#17: start worker process 24
nginx.1     | 2025/02/08 14:22:09 [notice] 17#17: start worker process 25
nginx.1     | 2025/02/08 14:22:09 [notice] 17#17: start worker process 26
<snip...>
dockergen.1 | 2025/02/08 14:22:09 Generated '/etc/nginx/conf.d/default.conf' from 4 containers
dockergen.1 | 2025/02/08 14:22:09 Running 'nginx -s reload'
nginx.1     | 2025/02/08 14:22:09 [notice] 17#17: signal 1 (SIGHUP) received from 57, reconfiguring
nginx.1     | 2025/02/08 14:22:09 [notice] 17#17: reconfiguring
dockergen.1 | 2025/02/08 14:22:10 Watching docker events
nginx.1     | 2025/02/08 14:22:09 [notice] 17#17: using the "epoll" event method

For reference, here is the dockerfile that I am using, the two changes are to set the VIRTUAL_HOST and the location of the db.env file with an appropriately updated db.env file:

services:
  db:
    image: mariadb:10.11
    command: --transaction-isolation=READ-COMMITTED
    restart: always
    volumes:
      - db:/var/lib/mysql:Z
    environment:
      - MYSQL_ROOT_PASSWORD=12345
      - MARIADB_AUTO_UPGRADE=1
      - MARIADB_DISABLE_UPGRADE_BACKUP=1
    env_file:
      - /path/to/your/db.env

  redis:
    image: redis:alpine
    restart: always

  app:
    image: nextcloud:apache
    restart: always
    volumes:
      - nextcloud:/var/www/html:z
      # NOTE: The `volumes` config of the `cron` and `app` containers must match
    environment:
      - VIRTUAL_HOST=servhostname.local
      - LETSENCRYPT_HOST=
      - LETSENCRYPT_EMAIL=
      - MYSQL_HOST=db
      - REDIS_HOST=redis
    env_file:
      - /path/to/your/db.env
    depends_on:
      - db
      - redis
      # Added proxy container dependency below. 
      # It is unclear on when or why it happens, but sometimes NC manages to start before the proxy 
      #  and it breaks for whatever weird reason resulting in the need of manual proxy container restart.
      - proxy
    networks:
      - proxy-tier
      - default

  cron:
    image: nextcloud:apache
    restart: always
    volumes:
      - nextcloud:/var/www/html:z
      # NOTE: The `volumes` config of the `cron` and `app` containers must match
    entrypoint: /cron.sh
    depends_on:
      - db
      - redis

#  letsencrypt-companion:
#    image: nginxproxy/acme-companion
#    restart: always
#    environment:
#      - DEFAULT_EMAIL=
#    volumes:
#      - certs:/etc/nginx/certs:z
#      - acme:/etc/acme.sh:z
#      - vhost.d:/etc/nginx/vhost.d:z
#      - html:/usr/share/nginx/html:z
#      - /var/run/docker.sock:/var/run/docker.sock:z,ro
#    networks:
#      - proxy-tier
#    depends_on:
#      - proxy

# self signed
  omgwtfssl:
    image: paulczar/omgwtfssl
    restart: "no"
    volumes:
      - certs:/certs
    environment:
      - SSL_SUBJECT=servhostname.local
      - [email protected]
      - SSL_KEY=/certs/servhostname.local.key
      - SSL_CSR=/certs/servhostname.local.csr
      - SSL_CERT=/certs/servhostname.local.crt
    networks:
      - proxy-tier

volumes:
  db:
  nextcloud:
  certs:
  acme:
  vhost.d:
  html:
  dhparam:

networks:
  proxy-tier:

I believe this example should be configured so that nginx-proxy produces more granular log messages, or at least put a comment in the compose.yaml that provides an environment variable that can be set to increase nginx verbosity. I'm not sure how to fix the example so that it actually runs with the self-signed certificate, though.