Run nodebb inside of docker container

[pwFoo]

Hi,
I tried to run nodebb/docker with a redis db as swarm / stack.

compose:

version: '3.7'
services:

  nodebb:
    image: nodebb/docker
    networks:
      webproxy:
      redis:
    labels:
      # caddy swarm reverse proxy with lets encrypt
      caddy.address: "nodebb.example.com"
      caddy.targetport: '4567'
      caddy.log: '/ stdout "[{host}] {remote} - {user} [{when}] \"{method} {uri} {proto}\" {status} {size}"'
      caddy.errors: "stdout"  
    stdin_open: true
    tty: true         
    volumes:
      - build:/usr/src/app/build
      - uploads:/usr/src/app/public/uploads
    deploy:            
      restart_policy:  
        condition: any 
        delay: 10s     
        max_attempts: 5
        window: 60s

  redis:
    image: redis:5.0.7
    networks:
      redis:
    volumes:
      - data:/data
    command: 'redis-server --appendonly yes'  
    

networks:
  webproxy:
    external:
      name: proxy_webproxy
  redis:

volumes:
  data:
  build:
  uploads:

After setup is finished I click the "Launch" button and that kills the container:

2019-12-28T11:54:49.017Z [19] - info: [build] Asset compilation successful. Completed in 140.239sec.

     

NodeBB Setup Completed. Run "./nodebb start" to manually start your NodeBB server.

Starting NodeBB
    "./nodebb stop" to stop the NodeBB server
    "./nodebb log" to view server output
    "./nodebb restart" to restart NodeBB

Would be no problem because of swarm restart_policy a new container should start...
But after that I get the installer again instead of the installed nodebb forum.

So something seems to be wrong. Redis DB is empty and nodebb volumes contains data.
See below:

rancher:/var/lib/docker/volumes# ls -lh nodebb_*/_data/
nodebb_build/_data/:
total 12
-rw-r--r--    1 root     root          11 Dec 28 11:54 cache-buster
drwxr-xr-x    2 root     root        4.0K Dec 28 11:33 export
drwxr-xr-x    7 root     root        4.0K Dec 28 11:53 public

nodebb_data/_data/:
total 132
-rw-r--r--    1 999      ping      129.8K Dec 28 11:52 appendonly.aof

nodebb_uploads/_data/:
total 20
drwxr-xr-x    2 root     root        4.0K Dec 28 11:25 category
drwxr-xr-x    2 root     root        4.0K Dec 28 11:25 files
drwxr-xr-x    2 root     root        4.0K Dec 28 11:25 profile
drwxr-xr-x    2 root     root        4.0K Dec 28 11:25 sounds

So how to install with persistent data?

Why no redis data is written to the volume?
I'm new with redis, but it should work without password because I haven't set an password?
Redis connect works fine from nodebb to redis container:

root@a8dd26787f8f:/usr/src/app# telnet redis 6379
Trying 10.0.19.2...
Connected to redis.
Escape character is '^]'.
quit
+OK
Connection closed by foreign host.

[pwFoo]

If I restart the running nodebb container it would work, but docker swarm respawn a new container to the first one after I manually restart it...

Also tested to add the install directory as volume if there would be changes after web install is finished, but same problem.
Each time a new container is spawned I get the web installer again...

It would help to now WHY I get the installer each time and how to solve that with the nodebb/docker image (with additional persistent volumes)?

[pwFoo]

@Staff
After some more debugging I found the config.json file inside of the "/usr/src/app" directory, which prevents the image to update the nodebb app by pulling a new image...

Any change to get the config.json moved to a sub directory "config"? Or wouldn't work nodebb inside of a docker container that way?

[fais3000]

I was facing this same issue. After a lot of hit and try this is the final docker-compose that i was able to make it work. Note I'm using Mongo instead of redis.