Startup fails with `unable to find /root/.my.cnf`

[m-strasser]

I'm trying to set up the wallabag docker container on a Raspberry Pi 3 running Raspbian.

I cloned the repository, built the image from the Dockerfile and exchanged the images in the docker-compose.yml accordingly (wallabag_rpi and rpi-mariadb):

version: '3'
services:
  wallabag:
    image: wallabag_rpi
    environment:
      - MYSQL_ROOT_PASSWORD=wallaroot
      - SYMFONY__ENV__DATABASE_DRIVER=pdo_mysql                                   
      - SYMFONY__ENV__DATABASE_HOST=db
      - SYMFONY__ENV__DATABASE_PORT=3306
      - SYMFONY__ENV__DATABASE_NAME=wallabag                                      
      - SYMFONY__ENV__DATABASE_USER=wallabag                                      
      - SYMFONY__ENV__DATABASE_PASSWORD=mydbpassword          
      - SYMFONY__ENV__DATABASE_CHARSET=utf8mb4                                    
      - SYMFONY__ENV__MAILER_HOST=127.0.0.1                                       
      - SYMFONY__ENV__MAILER_USER=~
      - SYMFONY__ENV__MAILER_PASSWORD=~
      - SYMFONY__ENV__FROM_EMAIL=wallabag@mydomain.com                          
      - SYMFONY__ENV__DOMAIN_NAME=wallabag.mydomain.com                        
    ports:
      - "8002:80"
    volumes:
      - wallabag:/var/www/wallabag/data
      - wallabag_images:/var/www/wallabag/web/assets/images                       
  db:
    image: monstrenyatko/rpi-mariadb
    environment:
      - MYSQL_ROOT_PASSWORD=wallaroot                      
    volumes:
      - /opt/wallabag/data:/var/lib/mysql                                         
volumes:
  wallabag:
  wallabag_images:

Starting a wallabag_rpi container then failes with:

Starting provisioner...
 [WARNING]: Found both group and host with same name: localhost
PLAY [localhost] bin dev entrypoint.sh etc home lib media mnt opt proc root run sbin srv sys tmp usr var 
TASK [Gathering Facts] bin dev entrypoint.sh etc home lib media mnt opt proc root run sbin srv sys tmp usr var ok: [localhost] 
TASK [needed dirs] bin dev entrypoint.sh etc home lib media mnt opt proc root run sbin srv sys tmp usr var ok: [localhost] => (item=/var/www/wallabag/app) ok: [localhost] => (item=/var/www/wallabag/app/config) ok: [localhost] => (item=/var/www/wallabag/data) ok: [localhost] => (item=/var/www/wallabag/data/assets) ok: [localhost] => (item=/var/www/wallabag/data/db) 
TASK [write parameters.yml] bin dev entrypoint.sh etc home lib media
mnt opt proc root run sbin srv sys tmp usr var ok: [localhost] 
TASK [stat] bin dev entrypoint.sh etc home lib media mnt opt proc root run sbin srv sys tmp usr var skipping: [localhost] 
TASK [notify install for sqlite] bin dev entrypoint.sh etc home lib media mnt opt proc root run sbin srv sys tmp usr var skipping: [localhost] 
TASK [wait for db container] bin dev entrypoint.sh etc home lib media mnt opt proc root run sbin srv sys tmp usr var ok: [localhost] 
TASK [add mariadb db] bin dev entrypoint.sh etc home lib media mnt opt proc root run sbin srv sys tmp usr var fatal: [localhost]: 
FAILED! => {"changed": false, "msg": "unable to find /root/.my.cnf. Exception message: (1045, \"Access denied for user 'root'@'172.19.0.2' (using password: YES)\")"} to retry, use: --limit @/etc/ansible/entrypoint.retry 
PLAY RECAP bin dev entrypoint.sh etc home lib media mnt opt proc root run sbin srv sys tmp usr var localhost : ok=4 changed=0 unreachable=0 failed=1  
Provisioner finished.

[Zayon]

Hi, I had the same issue.

It's the mysql_db ansible module that fails
Try adding POPULATE_DATABASE=false environment variable.
If your database is not already populated, try the manually populate it from the container once the provisioner finish without errors.

Also I see that you have a volume mounted in /var/www/wallabag/data this may prevent git from cloning the wallabag sources in the /var/www/wallabag directory so check that you have everything in the /var/www/wallabag directory.

[nickyeoman]

I also had the same issue.

In the hopes of saving someone else time, the Ansible script that runs is located here:
entrypoint.yml

for some reason when Ansible tries to connect to the db host, mariadb closes the connetion:
[Warning] Aborted connection 8 to db: 'unconnected' user: 'unauthenticated' host: '192.168.0.4' (This connection closed normally without authentication)

even though when you connect to the container:
docker exec -it dcebf5dcbda1 sh

you can connect to the database just fine:
mysql -h $SYMFONY__ENV__DATABASE_HOST -u root -p$MYSQL_ROOT_PASSWORD

But if I manually run the playbook:
ansible-playbook -i /etc/ansible/hosts /etc/ansible/entrypoint.yml -c local "$@"

I get the warning:
[WARNING]: Found both group and host with same name: localhost

But the playbook successfully completes.

If it doesn't (or if you have defined the mariadb user and database name in your compose file) you can try the POPULATE_DATABASE=false and then run php bin/console wallabag:install --env=prod -n from the /var/www/wallabag directory.

I think (untested) if you ran ansible-playbook -i ${HOSTNAME}, /etc/ansible/entrypoint.yml -c local "$@" instead of using an inventory and there was a check for if {{ database_name }} exists - name: add mariadb db in the playbook, it might work a little better. But I'm not sure how to stop mariadb from closing the connection.

hope this helps.

[j0k3r]

Ansible has been replaced by shell script few days ago #307, so I'm closing that issue.